While fertility schedules continue to shift towards later ages, it remains unclear why women’s birth probabilities in their late thirties and early forties have not risen more substantially (Beaujouan and Neels, 2025). One possible explanation lies in the fluidity of fertility intentions: some women may lower their desired number of children in response to changing life circumstances, social norms or perceptions of the most appropriate time to have children (Cooke et al., 2012; Iacovou and Tavares, 2011; Sennott and Yeatman, 2012). Consequently, these women may not attempt to conceive at all at older ages. If this is the case, then the limited increase in women’s late birth probabilities reflects an adjustment of their fertility goals rather than a biological constraint. It also implies that in an ideal situation in which the reasons why women reverse their childbearing intentions become less important or disappear, there is still scope for additional conceptions and births at older ages.

A contrasting explanation points to biological limits as the main barrier to further increases in late fertility (Beaujouan, 2022; ESHRE Capri Workshop Group, 2005; Menken, 1985). If many women are actively trying to conceive but are unable to do so due to biological barriers, including those who turn to medically assisted reproduction (MAR) without success, then the potential for further increases in late fertility is much more limited. Distinguishing between these explanations is essential, as each has distinct demographic and policy implications.

While the biological explanation for limited fertility recuperation at older ages has been developed extensively in previous research (Beaujouan, 2023; Ciganda and Todd, 2024; Leridon, 2017; Te Velde et al., 2012), the explanation that relates to the adaptive nature of fertility intentions has been less clearly articulated. Psychosocial adaptation processes play a crucial part in how women respond to the challenges of delayed childbearing (Kearney and White, 2016; Lei et al., 2025; Li et al., 2025). Some women deliberately revise their plans, while others make more unconscious adjustments in response to prolonged obstacles or their perception of whether they are able to have a child (Cooke et al., 2012; Wagner et al., 2019). After a long period of postponement, a once firm intention to have a child can evolve into uncertainty or be replaced by an intention not to have (more) children (Martin, 2017, 2021; Singh, 2026). These adjustments may serve as an unconscious or conscious coping strategy adopted to reduce emotional dissonance, but they may also reflect a voluntary re-evaluation of fertility goals over time (Koert and Daniluk, 2017; Martin, 2017). Fertility intentions and behaviours thus appear to be fluid and responsive to internal reassessment, life course timing and the external conditions that shape individuals’ choices. This emotional and cognitive flexibility is a defining feature of fertility behaviour and is consistent with research indicating substantial ambivalence in women’s expressions of fertility intentions (Bernardi et al., 2015; Bhrolcháin and Beaujouan, 2018; Jones, 2017; Trinitapoli and Yeatman, 2017; Weitzman et al., 2017).

Several types of constraints may affect women’s intentions to have children and attempts to conceive across the life course. Individuals who postpone having a child despite intending to do so soon may be experiencing situational constraints on childbearing, which are often linked to financial, health or partnership considerations (Cooke et al., 2012; Koert and Daniluk, 2017). As people get older, further barriers to having children may emerge that call their intentions into question (Berrington, 2021). Although societal acceptance of late childbearing is increasing as it becomes more common, driven by evolving attitudes and social contagion (Balbo and Barban, 2020; Lazzari et al., 2025), most people still perceive an upper age limit for childbearing across Europe. Women trying to conceive in their late thirties or forties may experience social stigma and feel “off time” in relation to the social deadlines for starting a family (Billari et al., 2011; Settersten Jr and Hägestad, 1996; Van Bavel and Nitsche, 2013).

Biological factors further intensify the challenges of fertility recuperation in cases of delayed childbearing. As women age, the time needed to conceive tends to increase and their probability of having a successful childbirth declines (Gnoth et al., 2003; Konishi et al., 2021; van Noord-Zaadstra et al., 1991). These biological restrictions impose a time limit on women’s childbearing, influencing not only their chances of conceiving, but also possibly their intentions and behaviours in the later stages of their reproductive lives. As women experience longer and potentially unsuccessful attempts to conceive at older ages, some may downplay their earlier intentions or begin to express uncertainty – not necessarily because they no longer wish to have children or are no longer trying to conceive, but as a way of coping with the emotional burden of unfulfilled goals, particularly when pregnancy feels increasingly out of reach (Badolato et al., 2025; Greil et al., 2010; Lazzari, 2025; Schmidt, 2009). While ART have expanded the window of reproductive opportunity for some women, these treatment options can only partially compensate for the biological constraints of ageing (Leridon, 2004). ART success rates remain largely dependent on ovarian quality, declining sharply after age 35 and reaching minimal efficacy in the forties (Delbaere et al., 2020; O’Brien et al., 2017). For women experiencing infertility, repeated unsuccessful attempts to conceive may lead them to become resigned to the idea that their goal of having a child is no longer attainable.

At the same time, women’s awareness of their declining ability to reproduce may create a sense of urgency around fertility decisions (Wagner et al., 2019). For some women, this urgency may result in greater efforts to conceive at later ages, despite the diminishing chances of success. While a growing number of women perceive themselves as infertile as they age (Passet-Wittig et al., 2020), it also appears that women who experience reproductive issues only once or twice are no less likely to have children than other women (Greil et al., 2024). This points to a strong commitment in women’s reproductive attempts, but their efforts to conceive may still be accompanied by heightened psychological uncertainty and emotional ambivalence linked to approaching age-related deadlines and longer duration of attempts. Therefore, opposing forces – the motivation to achieve parenthood and the uncertainty of success – may coexist and complicate the relationship between women’s stated goals and their actual behaviours.

The behavioural link between fertility intentions and childbirth, namely the act of trying to conceive, remains underexplored. This action reflects not only a strong fertility intention, but also the commitment to act on it. In his in-depth research into human reproductive motivations and behaviours, Warren B. Miller has identified a motivational and reproductive sequence, the Traits-Desires-Intentions-Behaviour framework (Miller, 1994). Of particular interest here is the way in which the sequence links fertility desires and intentions with fertility outcomes through instrumental behaviours (Miller, 1986). These behaviours encompass actions ranging from preventing to favouring conception. Preventing conception involves actions such as using contraception or abstaining from sexual intercourse, while favouring conception involves a “proceptive behaviour” that includes any action taken in favour of childbearing. Miller identified sexual and contraceptive practices, an adapted lifestyle, biological factors, perceived difficulty conceiving and seeking help to conceive as antecedents of conception. Studying the act of trying to conceive, its duration and its relationship to fertility intentions may provide further insight into proceptive behaviour.

Nevertheless, classifying conception-oriented actions is not straightforward. Different behavioural states relating to reproduction have been identified, ranging from “trying to get pregnant” to “being okay either way” through to inconsistent contraceptive use (McQuillan et al., 2011; Passet-Wittig and Lück, 2025; Rendall et al., 2020; Weitzman et al., 2017). Like intentions, the act of trying (or not trying) to conceive is presumably a complex process that is sensitive to age and is shaped by both opportunity and constraint (Guzzo et al., 2025). This process may become longer if women encounter reproductive issues and be stopped or reinforced via assisted reproduction if women begin to suspect that they are infertile.

By examining fertility intentions and attempts to conceive simultaneously, we can investigate how women’s behavioural efforts align with their stated fertility goals at different reproductive stages. Including information on the duration of women’s attempts and their perceived fertility status allows us to capture the interplay between biological realities and related subjective perceptions of reproduction. It also enables us to account for length bias (Basso et al., 2000; Zelen, 2004), whereby women who have been trying to conceive for a longer period of time are overrepresented among those who are currently trying – a phenomenon that is particularly prevalent among older women (Stevenson et al., 2026).

All of the fertility processes described thus far, from deciding to have children to actually having them, are shaped not only by individual circumstances, but also by the broader context. Their timing notably depends on contextual factors such as cultural expectations, perceptions of economic security and institutional support (Badolato et al., 2024; Brinton, 2016; Rindfuss et al., 2003; Setz et al., 2025). These factors also influence both the desirability and feasibility of late childbearing, and they vary substantially across age groups and countries.

At later reproductive ages, two institutional dimensions could be particularly salient: social age norms surrounding parenthood and access to MAR. However, the existing literature on the relationship between social age norms and late fertility, and the relationship between legal frameworks for MAR and late fertility, does not identify a straightforward association between the prevalence of these two aspects and levels of late fertility. In countries at advanced stages of fertility postponement, such as those in Southern, Northern and Western Europe, shares of late births are relatively high, but this trend is not always accompanied by widespread or equitable access to assisted reproduction (Beaujouan, 2020; Compans, 2024). In some parts of Eastern Europe, where late fertility emerged later, MAR has been promoted as a pro-natalist incentive (Cook et al., 2023; Frejka et al., 2016; Spoorenberg and Skirbekk, 2025; Szalma and Pélyi, 2025), while in Latin America, birth profiles remain comparatively early and access to MAR is limited (Dyer et al., 2016; Lima et al., 2018).

As late childbearing becomes more common and gains social acceptance, later attempts to have children are becoming increasingly normalised, particularly in Northern and Western Europe. By contrast, attitudes towards late parenthood remain more restrictive in many Eastern European countries and in the Southern Cone, even though fertility postponement is also progressing in these regions (Pardo and Cabella, 2018; Rosero-Bixby et al., 2009; Sobotka and Beaujouan, 2017). Consequently, it is likely that larger proportions of women are trying to have children at later ages in Northern and Western Europe. Additionally, we expect that a higher proportion of women will be trying to conceive for an extended period in Northern and Western Europe, where women start trying later, than in regions at less advanced stages of postponement. As a result, women with prolonged conception attempts accumulate in the population at risk and are disproportionately represented in cross-sectional data, creating a length bias at older reproductive ages that reinforces the observed regional differences.

Because fertility decisions are embedded in social networks, women’s perceptions of their own reproductive prospects are also shaped by the experiences of their peers (Balbo and Barban, 2014; Bernardi, 2003). In countries where postponement is more established, the greater visibility and prevalence of age-related reproductive challenges may foster social learning about the limits of late fertility. This could prompt some women in these contexts to temper their intentions or to approach their pregnancy attempts with more ambivalence.

Our study of the intersection between fertility intentions and attempts to conceive enables us to observe regional specificities and determine whether certain aspects depend on the stage of fertility postponement. Fertility postponement is most advanced in Northern and Western Europe (starting in the 1970s), while it is less advanced in Eastern Europe (with a rapid start in the 1990s) and it has progressed slowly starting in the 2000s in South America (Spoorenberg and Skirbekk, 2025). Due to the complexity of the factors involved, this study will not attempt to identify the institutional factors at play in reproductive experiences across ages. Instead, we assume that the stage of fertility postponement is an adequate indicator of the perceived acceptability and feasibility of late fertility in each region.

We use the data from the second round of the Generations and Gender Surveys (GGS-II), which began in the 2020s (see Table 1). For this cross-sectional study, we rely on the first wave of Round 2, drawing on data from 14 participating countries: Denmark, Norway, Finland, Estonia, Austria, Germany, the Netherlands, France, the United Kingdom, Croatia, Czechia, Moldova, Buenos Aires (Argentina) and Uruguay.

For comparative clarity, countries are displayed or pooled according to their geographical region: •

Northern Europe: Denmark, Norway, Finland, Estonia

This classification, based on geographical proximity, also broadly reflects the differences in the timing and progression of fertility postponement. Previous research shows that the Nordic countries and most Western European countries experienced the earliest onset of fertility postponement, starting in the 1970s (Beaujouan, 2020). A few countries, including Austria and Croatia, entered this process in the 1980s, followed by Estonia and other Central and Eastern European countries in the 1990s. By contrast, postponement has only recently emerged in Latin America, primarily since the 2000s, and its progression has been slow (Spoorenberg and Skirbekk, 2025). Including the Southern Cone region provides a useful contrast between settings where fertility postponement is well established and those where it has started more recently, while also capturing variation across European countries at different stages of this process.

We utilise the information on short-term fertility intentions, self-perceived fertility status, pregnancy attempts, duration of attempts and current number of children (parity at the time of the survey).

Short-term intentions were measured with the question:

“Do you intend to have a/another child during the next three years? Please take into account only biological children”.

Response options included: “Definitely not”, “Probably not”, “Unsure”, “Probably yes”, “Definitely yes” and “Currently expecting a child”. In addition, respondents could refuse to answer (“Refusal”) or discontinue the survey before completing the question (“Incomplete survey”).

Perceived fertility status was measured with the question:

“Some people are not physically able to have children. As far as you know, is it physically possible for you, yourself, to have a/another baby?”

Available response options included: “Definitely not”, “Probably not”, “Unsure”, ¹ “Probably yes” and “Definitely yes”, alongside the non-substantive categories “Refusal” and “Incomplete survey”.

Pregnancy attempts were captured with the question:

“Are you or your current partner trying to get pregnant?”

Response categories were: “Yes”, “No”, “Don’t know”, “Refusal” and “Incomplete survey”. As individuals who perceived themselves as definitively infertile or who were sterilised were not asked this question about “trying to get pregnant”, they were assigned to a separate analytical category (“Infertile/Sterile”) to retain them in the dataset and accurately reflect the share who were excluded from this question due to perceived or actual biological limitations. ²

Finally, to determine how long respondents had been trying to conceive at the time of the interview, we used the information on when they had started trying, alongside the month and year of the interview. Based on this calculation, we created a four-category variable reflecting the duration of attempts to conceive: “Less than a year”, “1–2 years”, “More than 2 years” and “Unknown”. ³

Parity was calculated based on the number of biological children the respondents reported having at the time of the survey.

For analytical consistency, non-substantive responses (“Don’t know”, “Refusal”, “Incomplete survey”) were combined into a single category across all variables. The only exception was the duration of pregnancy attempts, as such cases were incorporated into the “Unknown” group.

To represent age-specific probabilities of first and second births by country, we use data from the HFD for the year closest to the survey year for each country in the GGS-II (see Table 1). These birth probabilities are not available for Germany, France, the United Kingdom, Moldova, Buenos Aires and Uruguay.

Our study focuses on women. This choice is informed by both data considerations and the analytical scope of the paper. Age- and parity-specific birth probabilities from the HFD are available only for women, and much of the existing research on fertility intentions and reproductive behaviour has similarly focused on women. Given the already broad comparative scope of the study, limiting the analysis to women helps us to maintain clarity and focus. We acknowledge that extending this framework to men would be valuable, and we consider this an important direction for future research.

We begin by providing an overview of fertility intentions, pregnancy attempts and birth probabilities across reproductive ages and by parity for each country. These figures are shown in the text for childless women and women with one child. The figure for parity 2+ is provided in the supplementary material (available online at https://doi.org/10.1553/p-535f-56a4). Using GGS-II data, we calculate the share of women reporting a definite short-term intention to have a child, as well as the share of women actively trying to conceive, by age group and parity for all 14 countries. Age-specific birth probabilities for first and second births are drawn from the HFD for the available eight countries (see Table 1). To enable cross-country comparisons, we present all indicators on a common age scale and arrange countries row-wise by region, allowing patterns to be compared both within and across regions.

As childbearing is increasingly postponed, uncertainty about future parenthood becomes a central feature of reproductive decision-making, particularly at later ages (Barker and Buber-Ennser, 2024; Singh, 2026; Trinitapoli and Yeatman, 2017). To capture this dimension, we then show the age-specific proportions of childless and parity-one women who are unsure about their short-term intentions to have a child across the full reproductive age range. Again, numbers for parity 2+ are shown in the supplementary material.

Having estimated the prevalence of these events, we then assess the correspondence between short-term fertility intentions and reproductive behaviour. The aim of this descriptive step is to explore whether intentions translate into action, and, conversely, whether women who are attempting to conceive also express a clear positive intention to have a child, thereby assessing the degree of alignment or misalignment between women’s intentions and behaviour. Given that we are focusing on women who report strongly intending or trying to have a child, we do this for all parities combined (zero, one and 2+).

First, we focus on women who answered “Definitely yes” to the question on whether they intend to have a child within three years, which is used here as the closest available proxy for an immediate and firm childbearing plan. This choice is grounded in prior research showing that disaggregating fertility intentions by degree of certainty sharpens their predictive value, and that definite positive short-term intentions are associated with substantially higher realisation rates (Barker and Buber-Ennser, 2024; Buber-Ennser et al., 2014). Among this group, we report the distribution of responses to the question about attempting to get pregnant across age groups and regional clusters to examine whether definite intentions are accompanied by active behaviour. We further decompose the group who report trying and intending to have a child by attempt duration categories, which allows us to assess whether the cumulation of attempts at older ages helps explain the variation by age in the proportion of women trying to have a child.

To complement this perspective, we then reverse the analytical logic. Here, the sample is restricted to women who report actively trying to conceive, and we examine the distribution of the responses regarding short-term intentions within this group. This approach allows us to assess whether the level of certainty about intending to have a child is consistently high among women attempting to get pregnant, and whether the strength of women’s intentions varies with age. To explore heterogeneity in intentions among women trying to get pregnant, we incorporate information on self-perceived fertility status and the duration of attempts to conceive. Since the question on attempts to get pregnant was not posed to women who reported definite self-perceived infertility, we examine positive and negative/uncertain responses ⁴ to the question on short-term intentions with respect to the perception of probable infertility, stratified by region.

Finally, to assess the independent associations of these key reproductive factors with negative or uncertain responses or a refusal to answer ⁵ the question on short-term fertility intentions (as opposed to positive intentions) among women attempting to conceive, we estimate a binary logistic regression model and report the associated odds ratios (OR) with 95% confidence intervals. In this model, we control for age in five-year groups, perceived infertility, duration of attempts, parity ⁶ and region.

All descriptive and multivariate analyses are conducted using survey weights provided in the GGS-II to ensure that the results are representative at the population level across countries.

Figures 1 and 2 present the age-specific patterns of three key indicators in 14 European countries: the proportion of women expressing strong short-term intentions to have a child within the next three years (red curve), the proportion of women actively trying to conceive (yellow curve) and the probability of having a child (black curve). Figure 1 shows these patterns for first births among childless women, while Figure 2 shows these patterns for second births among women with one child. The results for women with two or more children are shown in the supplementary material (Figure S.1). However, as would be expected given the strong two-child norm (Sobotka and Beaujouan, 2014), a very small proportion of these women report intending to have more children, which makes the differences difficult to interpret.

Three main observations arise from these data. First, as expected, probabilities for both first and second births begin to decline significantly among women aged 35–39 in the eight countries for which these data are available in the HFD. In these older age groups, the proportion of women who give birth is consistently lower than the proportion who are trying to conceive. Second, the proportion of women in this age group with strong short-term fertility intentions also decreases; however, in many countries, it remains relatively large until it declines sharply at ages 40–44. Third, the proportion of women who are actively trying to have a child follows a strikingly different pattern. At younger ages, a smaller share of women report trying to conceive at the time of the survey compared to both the proportion who report intending to have a child in the short term and the share who ultimately give birth at that age (yellow curve is below the other curves). However, at older ages, the reverse pattern is observed: the proportion of women who are trying to conceive is larger than the share of women who intend to have their first child within three years (the yellow curve is above the red curve), except in France, Czechia and Buenos Aires (Figure 1). The same pattern emerges for parity-one women (Figure 2). The regional contrast already visible among childless women becomes particularly pronounced at parity one: in the Nordic and the Central and Eastern European countries, the share of women attempting to become pregnant at age 40+ is relatively large, while it is much smaller in Western European countries and the Southern Cone region, and it is often closer to the share of women with short-term fertility intentions. The patterns observed in older age groups in most countries contradict the initial expectation that more women would have positive short-term fertility intentions than would try to have children at all ages.

Figures 3 and 4 focus specifically on uncertainty, showing the age-related patterns in the proportion of women who report feeling unsure about having a child within the next three years, presented separately for childless women and women with one child. Among childless women (Figure 3), uncertainty generally peaks in the mid- to late thirties in most Northern and Western European countries, as well as in the Southern Cone region. Among childless women in their forties, this level of uncertainty drops, with most reporting no intention of having children (this result is not shown). Among women with one child (Figure 4), uncertainty is likewise concentrated in the later reproductive years, but the largest proportions who report being uncertain are in their early or mid-thirties rather than their late thirties. Overall, uncertainty levels are particularly high in Western and Central European countries (Czechia and Croatia). Among low-parity women, uncertainty regarding short-term childbearing largely appears to be a feature of their later reproductive years, albeit with some differences by parity and region. These findings are in line with the expectations of Hypothesis 1.

Figure S.2 in the supplementary material illustrates the levels of uncertainty among women who have two or more children. The proportions of these women who report being uncertain decrease with age, as most women with two or more children simply express a negative intention in their late thirties (not shown).

The aggregate age patterns in fertility intentions and pregnancy attempts do not map clearly onto stages of fertility postponement, which contradicts Hypothesis 5. We expected to observe similar patterns in the Nordic and Western countries but instead find that women in the Nordic countries and the Central and Eastern European countries who have no children or one child only are more likely to report trying to conceive at older ages than their counterparts in Western countries and the Southern Cone region (see Figures 1 and 2). Nevertheless, this is an interesting pattern, and focusing on who intends to have children and who tries to conceive across different ages and parity levels may clarify these unexpected regional differences. It could also provide insight into the surprising phenomenon observed in most countries in Figures 1 and 2, whereby more women report trying to conceive than intending to have children after age 40.

Our analysis of women who strongly intend to have a child within three years reveals notable age-specific patterns regarding whether they are attempting to conceive (Figure 5). Although a significant proportion of women at every age report that they are not trying to get pregnant despite expressing definite short-term intentions to have a child, the percentage of women who indicate that they are currently trying to conceive increases with age across most regions. The Southern Cone is the only region that displays a distinct pattern: a comparatively large proportion of women report trying to have a child in their early twenties, while pregnancy attempts barely become more frequent with age among women with strong short-term intentions. Few women who indicate that they definitely intend to have a child in the short term state that they are infertile or sterile, except in the Southern Cone region, where the proportion can reach 10% in some age groups; there is no specific age pattern. In line with Hypothesis 5, the increase with age in the proportion of women who report trying to get pregnant among those with strong short-term intentions is particularly marked in regions with later fertility schedules, such as Northern and Western Europe.

These patterns suggest two possible dynamics. First, older women may act with greater urgency (i.e. transitioning faster from intention to action) as age becomes a critical factor. Second, there is likely an accumulation effect due to lower success rates at higher ages, with more women who have been trying to get pregnant for a longer period of time. Figure 6 decomposes the “trying” category in Figure 5 by duration of attempts. This figure shows that the duration of trying increases with age, supporting the interpretation that older age groups increasingly consist of a selective group of women engaged in longer-term attempts to conceive. It also shows that, within a region, the proportion of women who say they are “newly trying” (i.e. have been trying for less than a year) among those who report a strong intention to have a child does not increase with age. In some cases, this share even decreases, as can be seen in Western and Central and Eastern European countries after age 40. These findings lend more support to the accumulation hypothesis than to the acceleration hypothesis in Hypothesis 2.

Focusing on women who report trying to conceive across different age groups, Figure 7 reveals significantly more uncertainty and inconsistency in short-term fertility intentions among older women, particularly among those over age 40. The significant proportion of older women who do not express definitive positive intentions supports the first part of Hypothesis 3. This tendency appears to be more prevalent in Northern Europe and Central and Eastern Europe than in the Southern Cone region and Western Europe.

Women who perceive themselves as definitely physically unable to have a child are not asked whether they are trying to conceive. Therefore, this result cannot be attributed to women who perceive themselves as definitely infertile but are trying to conceive anyway. Nevertheless, it could be related to women who perceive themselves as probably infertile.

A significant proportion of women with inconsistent responses (i.e. those who report trying to get pregnant but provide negative/uncertain/refusal responses to the question on short-term fertility intentions) report having difficulty conceiving (19–27%, depending on the region, for the answer option “probably infertile”; see Table 2). This proportion is substantially higher than that among women who report that they intend to have a child in the short term and are actively trying to conceive (5–8%). These findings support the second part of Hypothesis 3, which states that perceived infertility is associated with inconsistent intentions. Table S1 in the supplementary material also shows that the proportion of women who report that they have been trying to conceive for more than two years steadily increases across age groups. This situation may lead women to become uncertain about the outcome, and thus to reconsider their fertility intentions.

At the same time, the majority of women who give inconsistent responses do not report having difficulty conceiving (inferred from columns 4 and 5 of Table 2). Therefore, perceived infertility does not appear to fully account for the increase in inconsistent responses at older ages. However, due to the varying lengths of time spent trying and the significant differences in fertility intentions and associated behaviours by age and region, it is challenging to draw conclusions. In the next section, therefore, we set up a regression model.

The regression model explores the main factors associated with trying to get pregnant but providing negative/uncertain responses regarding short-term fertility intentions (versus expressing positive intentions) (Table 3). ⁷ Perceived infertility is associated with a higher likelihood of providing inconsistent responses (OR = 2.77), and longer durations of attempts also increase this probability, particularly among women who report trying for more than two years (OR = 5.04). Women who report trying to have a child and already have two or more children are also the most likely to say that they do not intend to have a child or are unsure (OR = 2.75), while childless women are the least likely to say that they do not intend to have a child or are unsure.

The persistence of an age effect net of these factors indicates that discrepancies between intentions and behaviour at later ages cannot be attributed to physiological constraints alone. The odds of reporting negative or uncertain short-term intentions while trying to conceive display a clear J-shape according to age. Compared with women aged 25–29, they are higher at ages 20–24 (OR = 2.36) but rise much more sharply from the mid-thirties onwards, reaching their highest level at ages 40 and above (OR = 8.08). Results from the multivariate model support Hypothesis 4 by showing that the age gradient in inconsistent responses remains strong after accounting for perceived infertility, duration of attempts, parity and region. Moreover, it is likely that anticipation and uncertainty regarding life circumstances or the capacity to have children shape women’s fertility intentions in their later reproductive years.

Clear regional differences also persist net of reproductive characteristics. Women in Western Europe and the Southern Cone region are significantly less likely to give inconsistent responses than women in Northern Europe and Central and Eastern Europe. We observe that contradictory responses are not more prevalent in settings where fertility postponement is more advanced. Thus, these findings underscore the importance of considering contextual influences that are not simply aligned with the degree of fertility postponement.

We identified marked age-related patterns in fertility intentions and attempts to conceive. A significant proportion of women in their late thirties with no children or only one child say they intend to have a child in the short term (Figures 1 and 2). At those ages, women also express substantial uncertainty regarding their intentions (Figures 3 and 4). In contrast, the prevalence of strong positive fertility intentions is very low among women in their early forties, regardless of parity. These results are consistent with previous studies indicating that women in their mid- to late thirties often change their answers to questions about their fertility intentions from positive to uncertain or negative (Barker and Buber-Ennser, 2024; Gemmill, 2019; Kuhnt et al., 2021). Such transitions are sometimes prompted by shifts in personal circumstances, such as changes in partnership status (Singh, 2026), or by the anticipation of having difficulty conceiving or the perception of being too old to have children (Martin, 2017, Wagner et al., 2019). Our study suggests that for a small but significant group of women, very long durations of trying to conceive at older ages (Table S.1) may contribute to these shifts from positive to less certain or negative intentions. Studies based on longitudinal rather than cross-sectional data could provide further insight into this question.

The percentage of older women trying to conceive is often higher than the percentage of women who definitely intend to have children, whereas the opposite is the case for younger women (Figures 1 and 2). Examining fertility intentions alongside pregnancy attempts explains this pattern. First, women with a strong intention to have children are more likely to be trying to conceive at older ages (Figure 5) and have often been trying for a long time (i.e. more than one year; Figure 6). Second, a greater proportion of women who are trying to conceive no longer report a clear intention to have a child from their late thirties onwards (Figure 7). This pattern indicates rising ambiguity about having children as women approach less fertile ages, which is consistent with women’s higher levels of uncertainty about their short-term fertility intentions just before they reach these ages (Figures 3 and 4) and the longer duration of their attempts to conceive at later ages (Figure 6). This ambiguity may be due to various factors, such as growing awareness of time limitations, internalised age norms regarding childbearing or the psychological toll of repeated unsuccessful attempts to conceive, rather than a genuine ambiguity about the desire to have a child (Gameiro and Finnigan, 2017; Liefbroer and Billari, 2010; Maheshwari et al., 2008; White and McQuillan, 2006).

Compared to their younger counterparts, women in their late thirties and forties are more likely to report being unsure or not intending to have a child while trying to get pregnant (Figure 7). We hypothesised that this pattern might be linked to the age-related increases in infertility and the longer duration of pregnancy attempts among older women. To test this hypothesis, we examined whether these biological factors account for the age-related variation in inconsistent responses. However, even after controlling for perceived infertility, duration of attempts, parity and region, a strong age gradient in contradictory intentions persisted among women trying to conceive. This indicates that the experience or perception of reproductive difficulties alone may not explain the coexistence of motivation (manifested through active attempts to conceive) and uncertainty (manifested through an absence of strong positive intentions) at later reproductive ages. This suggests that even if they are not experiencing reproductive issues, some people who are actively trying to conceive may anticipate the possibility of not having children as their fecundity declines with age. They may then base their answers to questions about their fertility intentions on this perception. There may also be social or individual reasons for not expressing a clear intention to have a child while trying to become pregnant at older ages, such as perceived age norms or health issues. Survey questions that directly ask about these reasons would help to shed more light on this ambivalence.

A much larger share of women with strong short-term intentions to have a child report attempting to do so after age 35 in Northern and Western European countries, which are at more advanced stages of fertility postponement, than in Central and Eastern European countries and the Southern Cone region (Figure 5). These differences between countries are partly driven by the high proportion of women who have been trying to conceive for over one year by age 35 (Figure 6). The larger proportion of older women who report trying to conceive compared to those who express definite short-term intentions, and the concentration of long-term attempts in the older age groups, may indicate that the Nordic and Western countries have reached a stage of postponement in which more women experience reproductive difficulties. The prevalence of partnerships, access to assisted reproduction and other country-specific variations in exposure to childbearing may also contribute to differences between countries in attempts to conceive among women with a strong intention to have a child. Further research is needed to clarify the mechanisms involved, but these patterns raise broader questions about the limits of postponing births and the potential implications for future fertility levels.

Unexpectedly, we find that the gap at older ages between the proportion of women attempting to conceive and the proportion of women expressing definite positive short-term intentions is more pronounced in Northern European countries and Central and Eastern European countries than in Western European countries and the Southern Cone region (Figures 1 and 2). This pattern suggests a stronger tendency in the former group of countries for women in their late thirties and forties to try to get pregnant without expressing clear short-term intentions to have a child; a pattern that is also visible in Figure 7. Consistent with these descriptive patterns, the regression analysis confirms that among women who are trying to get pregnant in these countries, the share of those who report being uncertain or having no intention to have a child is larger, although the Central and Eastern European countries are at an earlier stage of the postponement transition than the Western European countries. These patterns suggest that there are regional differences in how uncertainty is translated (or not translated) into behaviour. In the Western European countries and the Southern Cone region, women who are unsure about having a child may be more likely to postpone or refrain from attempting to conceive, whereas in Northern and Central and Eastern European countries, women may continue trying to get pregnant despite having ambivalent or uncertain childbearing intentions. This also suggests that there are different normative or behavioural orientations towards pregnancy attempts, such as greater acceptance of trying “just in case” or being open to pregnancy without having fully committed intentions. The clear regional gradient further suggests that the contradictory responses may stem from variations in the understanding and reporting of fertility intentions and pregnancy attempts in different cultural contexts, including different interpretations of survey questions. Cultural differences in social norms and expectations surrounding late childbearing, such as misperceptions around medically assisted reproduction (Compans et al., 2026), may also shape how women reconcile their desires, expectations and actions. These patterns underscore the inherent social nature of reproduction (Almeling, 2015) and underline the need for qualitative research to improve our understanding of how reproductive intentions and behaviours are perceived and negotiated across different regions.

This study has enabled us to identify several complexities in real-life reproductive decision-making, particularly in the relationship between fertility intentions and pregnancy attempts. As noted earlier, fertility intentions are inherently fluid and situation-dependent, and their meaning likely changes with personal circumstances. Our findings confirm that “trying to get pregnant” is a complex, age- and context-sensitive process that requires careful conceptualisation and measurement. It is socially constructed, and previous research shows that pregnancy planning and attempts are shaped by social, religious and cultural norms (McQuillan et al., 2011; Passet-Wittig et al., 2020).

Our results further suggest that pregnancy attempts are difficult to capture with a single survey item, as respondents may interpret the concept differently across contexts and depending on their situation (Klerman, 2000; McQuillan et al., 2011; Sweeny et al., 2015). For instance, among older women who believe they are unlikely to naturally conceive or who have been trying to conceive for an extended period, it is unclear whether having unprotected intercourse is still perceived as “trying”, or whether only active engagement with fertility treatments qualifies as a pregnancy attempt. Such ambiguities may partly explain the contradictory patterns we observe between intentions and behaviour.

Our results reflect the fact that older women who have spent a longer period of time trying to conceive are disproportionately represented at higher ages, as they are more likely to be observed (in their first, second or subsequent years of trying). In our case, this length bias results in a larger proportion of women being observed trying to have a child at older ages than at younger ages. Controlling for attempt duration was thus essential to producing unbiased comparisons of reproductive behaviour by age and across countries.

Despite these measurement concerns, the measures related to pregnancy attempts in the Generations and Gender Surveys shed light on how conception behaviours unfold during the reproductive process, a behavioural stage that is commonly assumed to exist, yet is rarely observed. This allows researchers to move beyond a direct intentions-to-outcomes framework and analyse how intentions are, or are not, converted into tangible actions amid varying structural and biological constraints. Examining pregnancy attempts alongside intentions explicitly provides empirical insight into the instrumental behaviours that link intentions to outcomes in the motivational sequence described by Miller (1994).

Notably, our study re-emphasises that women’s intentions rarely lead to birth in a linear manner when they reach their late thirties and forties. At this life stage, and in all the countries under study, prolonged attempt durations and perceived difficulties in conceiving become more prevalent. Half of women aged 40 and over in Northern Europe and around 40% of women in this age group in Western Europe who report definitely intending to have a child within three years indicate that they have been trying to conceive for one year or more. Such situations can lead to more complex pathways to childbirth, causing individuals to question their intention to have a child. Those who decide to proceed may need to adopt more proactive conception behaviours, such as seeking help, using assisted reproduction (with or without donor gametes) or deciding to adopt. Others may adapt by choosing not to have children, which will have a retroactive (negative) effect on intentions to have a child. This retroactive effect must be taken into account in research frameworks on fertility goals and unrealised fertility (Beringer and Milewski, 2024; Broussard, 2026; Yeatman and Smith-Greenaway, 2025).

Building on this foundation, future research would benefit from more refined indicators of reproductive behaviour that capture the intensity and nature of conception efforts. Integrating information on sexual activity, ovulation tracking, timing relative to fertility windows and engagement with ART would allow for more precise assessments of conception probabilities, particularly at the population level. As delayed parenthood becomes increasingly common, incorporating richer measures on reproductive behaviour into large-scale surveys will be essential for capturing how individuals interpret and enact pregnancy attempts, and how these processes interact with fertility desires and intentions, uncertainty about having children and biological limitations across the life course.

Studying the relationship between fertility intentions and attempts to conceive provided valuable insights into how women adapt to the constraints and realities of reproductive ageing in the context of delayed parenthood. The sharp rise in uncertain fertility intentions and conflicting responses with age suggests that declining positive intentions partly reflect anticipatory adjustment to biological constraints, which in turn may reduce the likelihood of initiating or sustaining pregnancy attempts. For many women, scaling back their intentions may serve as a coping mechanism that helps them reconcile their reproductive aspirations with their perceived limits. Our results show that women in their late thirties and forties increasingly experience prolonged and difficult periods of attempting to conceive, highlighting the tension between their continued motivation and their growing uncertainty about their ability to achieve this goal. This tension may become even more pronounced as fertility-related behaviours increasingly involve time-intensive, costly and medically mediated processes, particularly with the growing reliance on ART at higher ages (Adamson et al., 2023; Fauser et al., 2024). In such circumstances, acting on fertility intentions requires greater resources, planning and persistence.

Creating social and structural conditions that enable women to fulfil their childbearing intentions at an earlier stage of adulthood, when the desire for children and the biological capacity to conceive are more likely to coincide, could help to reduce the risk of severe reproductive setbacks. At the same time, it is necessary to support women who wish to become parents later in life, given the significant proportion of women trying to do so. Designing such support is challenging, however, because strong aspirations to have a child often coexist with uncertainty about their realisation.