Hyperinsulinemia and insulin-resistance are metabolic disturbances associated with obesity, essential hypertension, hypertriglyceridemia, glucose intolerance, overt non-insulin dependent diabetes mellitus, polymetabolic syndrome and atherosclerotic disease. The assessment of in vivo insulin sensitivity (SAI in vivo) changes achieved by life style modifications or drug interventions require a reproducible technique. To evaluate the day-to-day intra-individual repeatability of SAI-in vivo, we determined the variation in the SI index (calculated from the Minimal Model of Bergman modified by insulin or MMins) in 11 subjects with a wide range of insulin-resistance. SI (first study) varied from 0.82 to 8.48 x 10(-4) min-1/microU.mL (4.43 +/- 2.85 x 10(-4) min-1/microU.mL mean +/- SD) and highly correlated with SI (second study) (r = 0.89; p = 0.0002). The average interday coefficient of variation was 20.9 +/- 13.9% and was similar in subjects with low or high SI values. We also measured SAI in vivo by assessing the rate of serum glucose decline induced by human cristalline insulin 0.025 U/kg IV dose after a 12-14 hours fasting period (a modified Bonora's method or BBD) in 11 subjects. No subject presented biochemical or symptomatic hypoglycemia. SAI in vivo values determined by BBD varied from 21 a 234 mumol/ml/min (134 +/- 64.8 mumol/ml/min, mean +/- SD). We found a highly significant correlation between SI values obtained from MMins and SAI in vivo assessed by the BBD (r = 0.89, p = 0.0002). Our results suggest that the Mmins is a fairly reproducible procedure and that a BBD is an acceptable option to quantify SAI in vivo, mainly when a fast-execution practice is necessary or cost restrictions are required.