We systematically reviewed English-language publications arising from prospective cohort studies of the association between coronary heart disease risk and body mass index. Eighty published articles from 46 studies were identified. Two thirds of the studies, including all 14 studies with at least 500 cases ("larger" studies), reported a positive or J-shaped association. Among these 14 larger studies, the average increase in coronary heart disease risk for each 2 kg/m (2) higher body mass index was 14%. There was only limited evidence of effect modification by age, sex, ethnicity, or other variables. Evidence from many studies, including randomized controlled trials, implies that the association between coronary heart disease risk and body mass index is partly (or perhaps wholly) mediated by high blood pressure, dyslipidemia, and impaired glucose tolerance.