The authors studied the prevalence of the metabolic syndrome in the 1999-2002 National Health and Nutrition Examination Survey (NHANES) according to the World Health Organization, National Cholesterol Education Program (NCEP), and International Diabetes Federation (IDF) definitions. There was 92.9% agreement between the NCEP and IDF definitions. The IDF prevalence was higher (p = 0.001) due to more men fulfilling its criteria than the NCEP's (39.9 +/- 1.7% vs. 33.6 +/- 1.6%; p = 0.007). If central obesity were not a prerequisite, the IDF prevalence would increase slightly to 40.3 +/- 1.1%. Subjects categorized as having the metabolic syndrome under IDF but not NCEP tended to be men, younger, and leaner. Their prevalence of self-reported coronary heart disease was not significantly different from that of other metabolic syndrome patients. Whether waist circumference is a prerequisite does not affect the diagnosis of the metabolic syndrome in the United States. The IDF definition identifies additional individuals at risk for cardiovascular disease.