Aims To compare the risk of sudden death and non-sudden death from myocardial infarction associated with the metabolic syndrome (MetS) in asymptomatic men. Methods and results The mortality status of 6678 middle-aged men from the Paris Prospective Study I, who were free of diabetes and coronary heart disease (CHD) at the baseline examination, has been investigated over 21 years of follow-up. The sagittal abdominal diameter was substituted for waist circumference, and HDL cholesterol was unavailable. The presence of three abnormalities and the presence of abdominal adiposity plus at least two abnormalities defined the MetS, using the NCEP-ATP III and IDF criteria, respectively. Frequency estimate of the MetS was 14.4 and 16.7%, using the NCEP-ATP III and IDF criteria, respectively. The MetS increased the risk of sudden death and non-sudden death by 68% [95% confidence interval (CI) 1.05-2.70] and 38% (95% CI 0.95-2.01), respectively, after adjustment for other CHD risk factors (P for the comparison of the hazard ratios = 0.25). Hazards ratio using the IDF criteria were 2.02 (95% CI 1.30-3.14) and 1.69 (95% CI 1.20-2.38), respectively, (P = 0.26). Conclusion In healthy middle-aged men, the MetS increased the risk of sudden death and, to a lesser extent, the risk of non-sudden death over 21 years independent of CHD risk factors.