We assessed the association between first-trimester vaginal bleeding and singleton infant outcomes in a hospital-based population of 11,444 nondiabetic women. Low birth weight (LBW), shortened gestation, LBW at term, and neonatal death occurred more often in women reporting first-trimester bleeding than in those who never bled. These relationships remained statistically significant after adjusting for confounding factors. Women who experienced vaginal bleeding limited to the first trimester (N = 1174) had double the risk of delivering a preterm infant compared with those experiencing no bleeding (adjusted risk ratio = 2.0; 95% confidence interval 1.6-2.5). Bleeding limited to the first trimester was associated with a 1.6-fold risk of delivery of a term LBW infant (95% confidence interval 1.3-2.0). These results suggest that first-trimester vaginal bleeding is an important correlate of adverse infant outcomes.