A clustering of insulin resistance, hypertension and dyslipidemia has been labeled as the metabolic syndrome. Asians have a lower frequency of obesity than do Caucasians, but have an increasing tendency toward metabolic syndrome. Most data on metabolic syndrome are based on studies from Western countries with only limited information derived from Asian populations. We conducted a cross-sectional study of individuals aged 30-60 yr in workplace settings. We examined and analyzed the health data of 1,384 Japanese, Koreans and Mongolians for metabolic syndrome based on the modified definitions of the working definition proposed by the Third Report of the National Cholesterol Educational Program Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (ATP III definition). The prevalence of metabolic syndrome using the ATP III-BMI30 and ATP III-BMI25 definitions was 7% and 12% for Japanese, 7% and 13% for Koreans, and 12% and 16% for Mongolians, respectively. With the exception of obesity, the prevalences of individual metabolic abnormalities within each of the three Asian groups were similar to each other and to reported rates of prevalence in the U.S.A. Nevertheless, the values of sensitivity and specificity by the metabolic syndrome definitions are remarkably different relative to ethnicity. A universal metabolic syndrome definition is inappropriate for comparisons of metabolic syndrome among Asian ethnic groups. We believe that the ATP III-BMI25 definition is suitable for the determination of metabolic syndrome among Japanese and Koreans, and that the ATP III-BMI30 is more appropriate for Mongolians.