Some techniques for the evaluation of insulin resistance (IR), such as the clamp technique, are not viable for the study of large populations; and for this reason, alternative approaches based on fasting plasma glucose (FPG) and plasma insulin (FPI) have been proposed. The aim of this study was to compare the IR calculations obtained from FPI and FPG values with the insulin sensitivity (IS) index derived from the minimal model. Eighty-seven healthy subjects with a wide range of body mass index (18 - 44 kg x m -2) and 16 DM2 non-obese patients were included in the study. All of the patients underwent a frequently sampled intravenous glucose tolerance test (FSIGTT), and the minimal model of glucose was used for the estimation of insulin sensitivity (IS MINIMAL ). The HOMA-IR index, the Avignon index, and the quotient FPG/FPI were used to calculate basal steady-state IR. The basal IR value that best correlated with IS was Log (1/HOMA-IR) (r = 0.70, p < 0.001). All of the basal indices showed a high correlation with each other. In conclusions, insulin sensitivity indices as determined from the basal glycaemia and insulinemia values are not good estimators for metabolic reality from the perspective of the minimal model. Nevertheless, they might well have an IR screening value for epidemiological studies, as long as there is no pancreatic beta-cell dysfunction.