To determine whether [2(3)H], [3(3)H], and [6(14)C]glucose provide an equivalent assessment of glucose turnover in insulin-dependent diabetes mellitus (IDDM) and nondiabetic man, glucose utilization rates were measured using a simultaneous infusion of these isotopes before and during hyperinsulinemic euglycemic clamps. In the nondiabetic subjects, glucose turnover rates determined with [6(14)C]glucose during insulin infusion were lower (P less than 0.02) than those determined with [2(3)H]glucose and higher (P less than 0.01) than those determined with [3(3)H]glucose. In IDDM, glucose turnover rates measured with [6(14)C]glucose during insulin infusion were lower (P less than 0.05) than those determined with [2(3)H]glucose, but were not different from those determined with [3(3)H]glucose. All three isotopes indicated the presence of insulin resistance. However, using [3(3)H]glucose led to the erroneous conclusion that glucose utilization was not significantly decreased at high insulin concentrations in the diabetic patients. [6(14)C] and [3(3)H]glucose but not [2(3)H]glucose indicated impairment in insulin-induced suppression of glucose production. These results indicate that tritiated isotopes do not necessarily equally reflect the pattern of glucose metabolism in diabetic and nondiabetic man.