In a nationwide study of Serbian births, in 2008, we estimated the influence of maternal prenatal smoking and environmental tobacco smoke (ETS) exposure on birth outcomes. Using stratified two-stage random cluster sampling, 2,721 women were interviewed in-person (response rates 98.1 %), and 2,613 singleton live births were included. Date of birth, gender, birthweight, birth height and head circumference were copied from the official hospital Birth Certificate. Six exposure categories were defined according to mother's smoking history and exposure to ETS. We calculated adjusted mean values and group differences by analysis of covariance, and adjusted odds ratios for the low birthweight (LBW < 2,500 g). Compared to the reference category (non-smoking, non-exposed to ETS) we observed birthweight reductions in infants whose mothers smoked continuously during the pregnancy and were exposed to ETS (-162.6 g) and whose mothers were not exposed to ETS (-173 g) (p = 0.000, and p = 0.003, respectively), as well as reduction in birth length (-1.01 and -1.06 cm; p = 0.003 and p = 0.000, respectively). Reduction in birthweight and birth length related to exposure categories was not linear. Adjusted OR for LBW was almost tripled for mothers who smoked over the entire pregnancy and were non-exposed to ETS (aOR 2.85; 95 % CI 1.46-5.08), and who were exposed to ETS (aOR 2.68; 95 % CI 1.15-6.25). Our results showed strong effects of smoking throughout the pregnancy on reduced birthweight, birth length and head circumference, and increased risk for LBW. We were not able to detect an effect for ETS exposure alone.