Educational expansion has been a key driver of fertility postponement in Europe. Given the growing share of individuals with a migration background, understanding how education shapes fertility across these population subgroups is becoming increasingly important. To date, the intersection between education, migration background and fertility has received limited attention. Using Belgian register data (2000–2011) and shared frailty models for recurrent events, we examine the association between education and first, second and third births among 1.5 (migrated before age 18) and second (children of immigrants) generation women from six origin groups alongside native Belgian women (women without a migration background). We investigate (i) how educational gradients vary across origin groups and generations and (ii) how fertility varies within educational levels across origin groups, while also considering (iii) the role of selective entry into parenthood. Results reveal striking differences in how education shapes fertility across origin groups and birth orders. For first births, patterns by level of education are similar across origin groups, particularly among the second generation. For second births, variation by level of education is highly group-specific: positive gradients emerge among certain European origin groups, while remaining absent among Turkish, Maghrebi and other non-European groups. For third births, educational gradients are neutral for all groups. Substantial variation across origin groups and generations persists after controlling for educational levels, indicating that factors beyond education continue to differentiate childbearing behaviour. Finally, groups with a migration background show larger unobserved heterogeneity in fertility hazards, highlighting that minority groups cannot be considered homogenous in terms of fertility, and underscoring the need to control for selection when comparing the education-fertility nexus across origin groups and generations.