The study is a part of an ongoing prospective cohort study on the relationship between the exposure to environmental factors during pregnancy and birth outcomes and health of newborns. We have measured personal PM2.5 level in the group of 407 non-smoking pregnant women during the 2nd trimester of pregnancy. On average, the participants from the city center were exposed to higher exposure than those from the outer city area (GM = 42.0 μg/m3, 95% CI: 36.8–48.0 vs. 35.8 μg/m3, 95% CI: 33.5–38.2 μg/m3). More than 20% of study subjects were affected by high level of PM2.5 pollution (above 65 μg/m3). PM2.5 concentrations were higher during the heating season (GM = 43.4 μg/m3, 95% CI: 40.1–46.9 μg/m3) compared to non-heating season (GM = 29.8 μg/m3, 95% CI: 27.5–32.2 μg/m3). Out of all potential outdoor air pollution sources (high traffic density, bus depot, waste incinerator, industry etc.) considered in the bivariate analysis, only the proximity of industrial plant showed significant impact on the personal exposure (GM = 54.3 μg/m3, 95% CI: 39.4–74.8 μg/m3) compared with corresponding figure for those who did not declare living near the industrial premises (GM = 36.2 μg/m3, 95% CI: 34.1–38.4 μg/m3). The subjects declaring high exposure to ETS (> 10 cigarettes daily) have shown very high level of personal exposure (GM = 88.8 μg/m3, 95% CI: 73.9–106.7 μg/m3) compared with lower ETS exposure (≤ 10 cigarettes) (GM = 46.3 μg/m3, 95% CI: 40.0–53.5 μg/m3) and no-ETS exposure group (GM = 33.9 μg/m3, 95% CI: 31.8–36.1 μg/m3). The contribution of the background ambient PM10 level was very strong determinant of the total personal exposure to PM2.5 and it explained about 31% of variance between the subjects followed by environmental tobacco smoke (10%), home heating by coal/wood stoves (2%), other types of heating (2%) and the industrial plant localization in the proximity of household (1%).