Introduction. The prevalence of metabolic syndrome depends on the criteria used for its classification. Three criteria in common use are those from International Diabetes Federation (IDF), the Adult Treatment Panel (ATP-III) or its update (ATP-IIIa).Objective. The prevalence statistic for generated for metabolic syndrome was compared for on the basis of each of the three criteria.Materials and methods. The sample consisted of 155 teachers and employees in the school of medicine. The average age was 40.9; 54.2% were men. The three criteria were applied and the prevalences were compared with the Wilcoxon test and Cohen's kappa. Results. Metabolic syndrome prevalence generated by each criterion was as follows: ATP-III was 12.3% (95%CI 7.5-18.5), ATP-IIIa was 34.8% (95%CI 27.4-42.9) and IDF 32.9% (95%CI 25.6--40.9). The prevalence indicated by ATP-III was lower than the ATP-IIIa or IDF prevalances (p < 0,001); however those of ATPIII-a and IDF were similar (p=0,083). Poor agreement was seen between ATP-III and ATP-IIIa (k=0.414, IC95% 0.409-0.420), and between ATP-III and IDF (k=0.374, IC95% 0.368-0.379); however, very good agreement was obtained between ATP-IIIa and IDF (k=0.957, IC95% 0.950-0.963).Conclusion. The new definitions for metabolic syndrome, ATP-IIIa and IDF, increase the prevalence statistic by three times. This occurred despite the inclusion in IDF of an obesity factor in the criteria set.