A cross-sectional population survey was carried out in 15 population groups (ethnicity includes Melanesian, Polynesian, Micronesian, Asian Indian and Chinese) in 9 developing countries: Fiji, Nauru, Kiribati, Cook Island, Niue, Western Samoa, New Caledonia, Mauritius and China (Beijing) in 1978-1987. The total sample included 4594 men and 4988 women aged 35-59 years. The aim of study is to report the prevalence of coronary heart disease (CHD) as indicated by ECG Minnesota coding, and risk factor levels and to describe the individual and ecological relationship between CHD prevalence and CHD risk factors among different ethnic groups in developing countries. Mauritians had the highest prevalence of CHD of these countries. Total serum cholesterol concentration and the prevalence of CHD were higher in Mauritius Chinese than in Beijing Chinese. Mean total cholesterol was lower than or equal to 5.2 mmol/l (200 mg/dl) in all population groups, except in Mauritians. Hypertensive subjects in most populations had a low cholesterol concentration. The prevalence of hypertension varied from 7 to 35% and mean body mass index (BMI) from 22.9 to 37.0 kg/m2. Smoking was more common in men (36-82%) than women (0.8-65%). Multiple logistic regression analysis using individuals as a unit of analysis showed that cholesterol and systolic blood pressure were significant independent predictors of CHD prevalence. When fasting or 2 hr post-load blood glucose was included in the model total cholesterol was no longer significant in men but remained significant in women. Ecological analysis using populations as units of analysis showed that the combination of several CHD risk factors could explain about 90% of the interpopulations variance of the CHD prevalence in women. The best models were those where 2 hr post-load glucose was included. Our study has demonstrated that the total cholesterol concentration of the population was consistent with the prevalence of CHD in the population. A considerable proportion of the variation in CHD prevalence across populations in developing countries can be explained by well-known risk factors. These data support the concept that retaining traditional balanced dietary habits and limiting salt intake together with avoiding smoking use are important activities for the prevention of cardiovascular disease (CVD) in developing countries.