OBJECTIVE: Comparison of BMI with waist circumference, waist-to-hip ratio (WHR), and waist-to-stature ratio (WSR) as a predictor of hypertension incidence. METHODS: A total of 1658 men and 1976 women of Mauritian Indian and Mauritian Creole ethnicity, aged 25-74 years, free of hypertension, diabetes, cardiovascular disease, and gout at baseline in 1987 or 1992, were re-examined in 1992 and/or 1998 using the same survey methodology. Hazard ratios (HRs) for hypertension incidence were estimated applying an interval censored survival analysis (R program) using age as timescale based on baseline obesity indicators. RESULTS: A total of 787 incident hypertension cases were identified during the follow-up. HRs for hypertension incidence adjusting for baseline systolic blood pressure and cohort corresponding to a 1 SD increase in BMI, waist circumference, WHR, and WSR were 1.20 (1.24), 1.19 (1.21), 1.14 (1.10), and 1.20 (1.26) in Mauritian Indian men (women) and 1.23 (1.32), 1.34 (1.23), 1.41 (1.13), and 1.43 (1.33) in Mauritian Creoles, respectively, indicating that all obesity indicators significantly predicted hypertension incidence except for WHR in Mauritian Creole women. Paired homogeneity tests showed that there was no difference between BMI and the other three indicators for most of the comparisons with two exceptions: WSR was stronger than BMI (P = 0.002) in Mauritian Creole men but BMI was stronger than WHR (P = 0.047) in Mauritian Indian women in predicting the incident cases of hypertension. CONCLUSION: The relation of the development of hypertension with BMI was as strong as that with central obesity indicators in the population studied.