We studied the impact of maternal exposure to environmental tobacco smoke (ETS) on birth weight (BW), low birth weight (LBW), and intrauterine growth retardation (IUGR) according to self-reported maternal smoking habits in a sample of 6,866 singleton births. We obtained data about parental characteristics and maternal active smoking (AS) and passive smoking at delivery via maternal questionnaires and medical records. We used three categories of smoking habits (nonsmokers and those who smoked 1-10 or >10 cigarettes per day) and defined ETS exposure as greater than or equal to 5 cigarettes per day smoked by others in the mother's presence. We used multiple regression and logistic regression procedures with adjustment for many associated covariates. We observed a significant reduction of the mean BW in infants of AS mothers. This reduction was only marginal for mothers who stopped smoking after recognizing their pregnancy. ETS exposure in 1,797 of 5,507 nonsmoking mothers reduced the mean BW of their infants by 53 g [95% confidence interval (CI), 24-82 g]. ETS exposure also significantly reduced BW in babies of AS mothers by 92 g (CI, 21-113 g) compared with BW of ETS-nonexposed AS mothers. The adjusted odds ratio (AOR) of LBW for ETS-exposed AS mothers was two times the LBW risk of ETS-nonexposed AS mothers(2.02; CI, 1.11-3.67); the AOR of ETS-exposed nonsmoking mothers was 1.51 (CI, 1.02-2.26). The AOR of IUGR for this group did not differ from unity (1.08; CI, 0.82-1.43). However, ETS exposure increased the AOR of IUGR for AS mothers from 1.64 (CI, 1.06-2.53) to 2.13 (CI, 1.70-2.67). ETS exposure reduced the BW of infants of nonsmoking mothers and contributed to additional BW reduction in infants of AS mothers. ETS exposure increased the risk of LBW but not that of IUGR in babies of nonsmoking mothers.