The significance of the metabolic syndrome (MetS) resides either in its ability to identify individuals at high risk of future disease and disability or in its ability to identify individuals in need of a specific treatment. We have previously shown that in the general population the ability of the MetS to identify individuals at increased risk of diabetes or cardiovascular disease (CVD) is inferior to the ability of established predicting models for these conditions. Although it may someday become routine to recommend treatment with insulin-sensitising agents for non-diabetic individuals with the MetS, most of whom are insulin resistant, there are no clinical trial data to support such a recommendation at the present time. Currently, the treatment of the MetS is based on treatment of its component parts. In the present paper, we examine the role of the MetS as defined by the National Cholesterol Education Program (NCEP) criteria in predicting all-cause and CVD mortality in patients with prevalent CVD from the San Antonio Heart Study (SAHS). This population contains a high proportion of Mexican Americans, who are at high risk of developing type 2 diabetes. After adjusting for age and gender, the MetS is moderately predictive of all-cause and CVD mortality. After further adjustment for diabetes, however, the effect of the MetS becomes non-significant in this population. Moreover, among non-diabetics with prevalent CVD, the MetS was not associated with either all-cause or CVD mortality. Thus, this study indicates that the effect of the MetS on these endpoints is primarily driven by the inclusion in the NCEP definition of diabetes, itself a well-established, potent CVD risk factor. In fact, the prevalence of diabetes in SAHS patients with CVD and the MetS was 42% compared with only 9% in patients with CVD, but without the MetS. This excess prevalence of diabetes appears to account for the enhanced all-cause and CVD mortality in subjects with the MetS. However, these results will need to be confirmed in other populations.