Objective: To assess the association between fetal macrosomia and adolescent obesity. Design: Longitudinal cohort study of the association between macrosomia and adolescent obesity. Subjects: Between 1 October 2005 and 1 February 2007, a follow-up study of live-born infants born in 1993-1995 in Wuxi, a suburban area of Shanghai, was conducted. Subjects with birth weight > 4000 g were selected as the exposed. For each exposed subject, one subject with a birth weight of 2500-4000 g, matched by year of birth, sex of infant, and type of institute at birth, was chosen as non-exposed. Clinical data were collected by structured interview and physical examination. Obesity was defined as body mass index (weight (kg)/height (m)) higher than the sex-age-specific criteria by the working group on obesity in China. Distribution of baseline characteristics and adolescent obesity rate between the exposed and non-exposed groups was compared. Results: A total of 1435 pairs of exposed and non-exposed subjects were included in the final analysis. No major difference in baseline characteristics (other than birth weight) was found between the exposed and non-exposed groups. Obesity rate was significantly higher in the exposed group (2.9%) than in the non-exposed group (1.6%). Adolescent obesity rates were 1.4, 1.9, 2.6, and 5.6%, respectively, in study subjects with a birth weight of 2500-3499, 3500-3999, 4000-4499, and ≥4500 g. The association between birth weight and adolescent obesity remained essentially the same when mother's demographic and anthropometric factors, breast feeding, and adolescent life-style factors were adjusted. Conclusion: Compared with infants of normal birth weight, infants with birth weight >4000 g, especially those >4500 g, are at increased risk of adolescent obesity. © 2009 Macmillan Publishers Limited All rights reserved.