To evaluate the impact of a first trimester induced abortion on the risks of low birth weight (LBW) and preterm birth in a subsequent pregnancy we conducted a pregnancy-based cohort study in Shanghai, China with recruitment from 15 general hospitals (or maternity and infant health institutes) from November 1993 to March 1998. Pregnant women with a history of induced abortion entered the abortion cohort and a reference cohort was established among women without such a history. All the subjects were enrolled before 64 days of gestation and interviewed five times until 42 days after delivery. Only singleton live births were analysed in this study. A total of 2953 pregnant women were enrolled and 2707 gave birth to live singletons. The overall incidence of LBW was 1.7%, 2.0% in the abortion cohort and 1.4% in the reference cohort. After controlling the potential confounders using logistic regression, the adjusted odds ratios were 1.7 (95% CI: 0.8-3.3) for LBW, 2.0 (95% CI: 0.9-4.7) for term LBW and 1.0 (95% CI: 0.6-1.5) for preterm birth, following an induced abortion pregnancy. Birth weight in the abortion cohort was 38.5 g higher than that of the reference cohort after adjustment for calendar year at recruitment, couples' occupation, education, age, infant sex, maternal body mass index at recruitment, contraceptive use and gestation age. Previous first trimester induced abortion did not significantly increase the risk of LBW or preterm birth. The study was performed in a low-risk population and results may not be applicable to other settings.