In this research, we calculate the probability and scale of population losses that may occur due to earthquake hazard in Bucharest. Losses are quantified in direct relation to the social vulnerability of people who are exposed to an earthquake event. Social vulnerability is based on index construction, using spatial decision rules to assess weights in a criteria tree using the SMCE-module of Ilwis software. To estimate building vulnerability, we used the Improved Displacement Coefficient analytical method in the SELENA software. For the earthquake loss estimation, we used the percentage of severely damageable residential buildings in each census unit (CU). The population loss estimation for the selected earthquake scenarios was obtained by multiplying the complex social vulnerability index with the estimated ratio of severely damageable buildings, for three selected earthquake scenarios and using the population numbers in each census unit. The maps represent the maximum affected population values, per census unit, in percentages. We provide useful estimates of the scale and severity of injuries, and link these with current levels of medical preparedness. In all scenarios, the CUs forming the Rahova neighbourhood revealed high loss values, due to significant problems in terms of the built environment and social vulnerability.