The association between socioeconomic position and health is generally believed to be weaker among women than men. However, gender differences in the relation between socioeconomic position and coronary heart disease have not been evaluated in a representative sample of the US population. The authors examined this association in the First National Health and Nutrition Examination Survey (1971-1993), a longitudinal, representative study of the US population (n = 6,913). Information on educational attainment, household income, and covariates was derived from the baseline interview, and that on incident coronary heart disease was obtained from hospital records/death certificates over 22 years of follow-up. Cox's proportional hazards models showed that education and income were inversely associated with incident coronary heart disease in age-only and multivariate models. Risk associated with education varied by gender (p = 0.01), with less than high school education associated with stronger risk of coronary heart disease in women (relative risk = 2.15, 95% confidence interval: 1.46, 3.17) than in men (relative risk = 1.58, 95% confidence interval: 1.18, 2.12) in age-adjusted models. Low education was associated with greater social and psychological risks for women than men; however, metabolic risks largely explained gender differences in the educational gradient in coronary heart disease.