OBJECTIVE - We have carried out international comparisons of the metabolic syndrome using the International Diabetes Federation (IDF) and National Cholesterol Education Program-Adult Treatment Panel III (ATP III) definitions. This analysis could help to discern the applicability of these definitions across populations. RESEARCH DESIGN AND METHODS - Nondiabetic subjects aged 35-64 years were eligible for analysis in population-based studies from San Antonio (Mexican Americans and non-Hispanic whites, n = 2,473), Mexico City (n = 1,990), Spain (n = 2,540), and Peru (n = 346). κ Statistics examined the agreement between metabolic syndrome definitions. RESULTS - Because of the lower cutoff points for elevated waist circumference, the IDF definition of the metabolic syndrome generated greater prevalence estimates than the ATP III definition. Prevalence difference between definitions was more significant in Mexican-origin and Peruvian men than in Europid men from San Antonio and Spain because the IDF definition required ethnic group-specific cutoff points for elevated waist circumference. ATP III and IDF definitions disagreed in the classification of 13-29% of men and 3-7% of women. In men, agreement between these definitions was 0.54 in Peru, 0.43 in Mexico City, 0.62 in San Antonio Mexican Americans, 0.69 in San Antonio non-Hispanic whites, and 0.64 in Spain. In women, agreement between definitions was 0.87, 0.89, 0.86, 0.87, and 0.93, respectively. CONCLUSIONS - The IDF definition of the metabolic syndrome generates greater prevalence estimates than the ATP III definition. Agreement between ATP III and IDF definitions was lower for men than for women in all populations and was relatively poor in men from Mexico City. © 2006 by the American Diabetes Association.