In this study we investigated a simple nonlabor-intensive method to evaluate insulin sensitivity and beta-cell function which is suitable for application in population studies. The method is a refinement of the modified Harano test and consists of a continuous low dose insulin (25 mU/kg.h) and glucose (4 mg/kg.min) infusion test (LDIGIT) lasting 150 min. Insulin sensitivity was evaluated as the MCR of glucose divided by the steady state serum insulin level achieved at the end of the test. Insulin secretion was expressed as the incremental area for C-peptide concentration during the first 15 min of the test. We compared the indices of insulin sensitivity and insulin secretion yielded by LDIGIT with those derived from the euglycemic clamp and the hyperglycemic clamp, respectively. Fifty-four subjects underwent a LDIGIT (33 with normal glucose tolerance and 21 with impaired glucose tolerance); of the 54, 19 were submitted to a euglycemic clamp, 18 to a hyperglycemic clamp, and 10 to a modified Harano test (insulin infusion, 50 mU/kg.h; glucose infusion, 6 mg/kg.min). LDIGIT overcame the drawbacks associated with the modified Harano test because it resulted in more stable final glucose levels and prevented the occurrence of hypoglycemic episodes. No significant differences were found between the insulin sensitivity index (ISI) of the LDIGIT and that of the euglycemic clamp for each group of subjects. Moreover, there was a strong correlation between the ISI determined by LDIGIT and the ISI determined by clamp (r = 0.90; P < 0.0001), and the best regression line was not different from the identity line, suggesting that the two indices are equivalent. The index of insulin secretion provided by LDIGIT correlated well with that of the hyperglycemic clamp (r = 0.82; P < 0.001) and was significantly higher in overweight subjects than in normal weight subjects. In conclusion, LDIGIT is a simple and accurate method to assess insulin sensitivity and secretion. It can be useful in population studies and in situations when more complex techniques are not feasible.