Objective: Determine effects of prenatal smoking and drinking on birthweight. Design: An ancillary study of a prospective, multinational cohort study (the Safe Passage Study) assessing fetal effects of alcohol. Setting: South African community of low socioeconomic status. Sample: Pregnant women with single, normal, liveborn infants (August 2007- January 2015). Methods: Prenatal alcohol exposure (PAE) was determined on multiple occasions during pregnancy and transformed into number of standard drinks (14 g pure alcohol) per day or binges (≥ 4 standard drinks per event). Cigarette exposure (PCE) was recorded from the number of days smoked per month and amount smoked on typical days. PAE and PCA, single or in combination, were expressed as continuous variables, exposure groups or trajectories, and compared with maternal and infant characteristics. Main outcome measures: BW z-scores (BWZs) and proportions of (severe) small-for-gestational age infants (SGA; < P10 (P3)) in relation to exposure metrics. Results: There were 4 098 participants (64.8% drinkers (63.3-66.3% CI), 63.0% smokers (61.4-64.34 CI), 45.8% dual users (44.3-47.3% CI); 739 SGA (18.0% (16.9-19.2%)) and 266 severe SGA (6.5% (5.8-7.3% CI)). BWZs were significantly lower with all exposure metrics, even low single use and when quitting during pregnancy. (Severe) SGA was increased with all exposure metrics, esp. with continued and/or dual exposure, even if low. Controlling for maternal variables which independently correlated with exposures and/or fetal growth, these findings persisted for exposure trajectories, and all dual exposure groups, but not for single exposure. Conclusions: Dual exposure to smoking and drinking (even low) reduces fetal growth.