Following up on original descriptions of clustering of cardiovascular risk factors (chiefly, glucose intolerance, dyslipidaemia and hypertension) around the presence of insulin resistance, the metabolic syndrome has recently been upgraded to the status of a disease entity with an inherent predictive value for cardiovascular disease. In pathophysiological terms, insulin resistance (of glucose metabolism) and the attendant compensatory hyperinsulinaemia are causally related to each of glucose intolerance, dyslipidaemia, high blood pressure and vascular dysfunction. The physiological mechanisms are concisely reviewed here. However, insulin resistance/hyperinsulinaemia alone is insufficient to cause these abnormalities, for which other pathogenic factors (e.g. ss-cell dysfunction for glucose intolerance) are required. The metabolic syndrome, on the other hand, has evolved from a set of statistical associations believed to carry an excess of cardiovascular risk. In the various existing definitions, a mixture of physical, metabolic and clinical variables have been used on grounds of predictive value or practical ease. These variables belong to different phenotypes, which are upstream, intermediate and proximal, respectively, in their relation to clinical disease. The resulting 'syndromes' usually lack a cogent conceptual structure, may reflect the particular data set from which they are extracted and may be of limited applicability. While overt diabetes, clinical hypertension and frank dyslipidaemia are often present together in the same patient, a subclinical syndrome with a distinct, probable aetiology and a proven power as a risk indicator remains to be identified.