OBJECTIVE The objective of the study was to evaluate whether maternal tobacco use is associated with an attenuation in fetal birthweight among women with nutritional and uteroplacental constraints. METHODS A population-based retrospective analysis of term (37 weeks or longer) singleton pregnancies delivered in Utah from 1991 to 2001. Birthweight (BW) and percent small for gestational age (SGA) (less than 10% for gestational age) among self-identified smokers and nonsmokers were compared. Adjusted odds ratios (ORs) were calculated to measure the association of maternal smoking with delivery of an SGA infant controlling for potential confounders across maternal strata. RESULTS Among the 424,912 gestations, 37,076 occurred in self-identified smokers. Mean BW was significantly less and the prevalence of SGA infants was significantly greater in tobacco-exposed infants across all maternal BMI strata (P < .001) as well as pregnancies complicated by diabetes (P < .001) and hypertensive disorders (P < .001). In a multivariable logistic regression model, tobacco exposure remained the significant associative factor for SGA (OR 3.53, 95% confidence interval 2.61 to 4.79) after selecting for the first birth in the study interval (n = 283,916). CONCLUSIONS Self-identified tobacco use increases the risk of a SGA infant at term across maternal strata.