The aim of this analysis was to examine the effect of environmental tobacco smoke exposure on the risk of small-for-gestational-age (SGA) birth. The study population included 2,283 nonsmokers from a nested cohort study undertaken in southern Connecticut from 1988 to 1992. The duration and intensity of exposures incurred at multiple locations during the third trimester of pregnancy were measured by postpartum interview. The effect of exposure on birth weight and on incidence of SGA birth was assessed by multivariate logistic and linear regression. An estimated 26.5% of the women had been exposed to environmental tobacco smoke for at least 1 hour per week during the third trimester. The median duration of exposure among the exposed over all locations was 5 hours per week. The adjusted odds ratio for SGA birth in exposed mothers compared with unexposed mothers, using a dichotomous exposure variable, was 0.82 (95% confidence interval: 0.51, 1.33). The adjusted birth weight difference associated with exposure was -1.2 g (95% confidence interval: -43.3, 41.0). No effect of environmental tobacco smoke exposure on fetal growth was seen in this relatively homogeneous upper middle class group of women exposed at low levels. This is reassuring for women exposed at low levels, but it does not exclude the possibility of an effect in women exposed to higher levels of environmental tobacco smoke.