Due to the association between insulin resistance (IR) and atherosclerosis, there is an interest in the development of techniques to evaluate insulin sensitivity (IS) in vivo. Fasting blood glucose, easy to use in study populations, has been used to evaluate IS and supplies a good evaluation of hepatic sensitivity, but not muscular sensitivity to insulin. HOMA is a mathematical model that predicts IS simply by measuring insulinemia and fasting blood glucose and shows good correlation with hyperinsulinemic-euglycemic clamp method, considered a gold standard in the measurement of IS. Thus, it has been shown a valuable alternative to the most sophisticated and difficult techniques in the evaluation of IR in humans. In our population, the cut value for the diagnosis of IR is Homa-IR higher than 2,71. QUICKI is another simple method, also based in the measurements of insulinemia and fasting blood glucose, that have good correlations with the metabolic syndrome markers, being able to discriminate satisfactorily different states of IR, in patients with different degrees of obesity and glucose tolerance. Direct methods of IS evaluation include insulin tolerance test (K ITT), insulin suppression test and hyperinsulinemic-euglycemic clamp technique that are described in this article. Hyperinsulinemic-euglycemic clamp technique supplies the best and purest information on the insulin action. Costs involved in its procedure, however, limit its use.