OBJECTIVE: To test association between overweight, central obesity and stature. DESIGN: Cross-sectional study carried-out between 1990 ± 1991. SUBJECTS: 951 adults (387 male and 564 female) aged 20 ± 64 y, resident in the metropolitan area of Sa Ä o Paulo, Brazil. MEASUREMENTS: Anthropometry, blood lipid concentrations (total, high density lipoprotein (HDL) and low density lipoprotein (LDL)-cholesterol, triglycerides (TGs) and blood glucose. Body mass index (BMI), waist-to-hip ratio (WHR) and waist circumference were used to identify overweight (BMI b 25 kgam 2 ), obdominal obesity (WHR tertile 3 and waist circumference tertile 3), respectively. The subjects were categorised as those of short stature (women`150 cm, men`162 cm) and those of normal stature (women !150 cm, men !162 cm). RESULTS: Prevalence of short stature was 19.6% and 15.4% in men and women, respectively. Short stature women had higher serum concentrations of total cholesterol, LDL-cholesterol, TGs and glucose than those of normal stature. Among men, this difference was not observed, except for glucose concentrations. Short stature women had high BMI and WHR means in some age categories, compared with those of normal stature. Both overweight and high WHR frequencies were greater in short stature women than in those of normal ones. In multivariate analysis, adjusted by age, income, marital status, education, physical activity and tobacco use, only women group with short stature compared with normal stature had signi®cantly risk of overweight an high WHR. In the same group there was no association with waist circumference. Among the men there was signi®cant opposite association with waist circumference. CONCLUSION: Short stature in women can potentially be an independent risk factor for overweight and high WHR.