Acceleration of glucose tolerance after repetitive intravenous glucose loads (Staub-Traugott effect) has not previously been tested in non-insulin-dependent diabetes (NIDDM). Six overt, untreated subjects were administered three hourly IV glucose tolerance tests (GTT). The glucose disappearance rate (K) changed very little between loads, indicating a suppressed Staub effect. However, insulin levels increased with each load. The characteristic loss of early phase insulin release after the first intravenous glucose injection was recovered with the third injection. After four months of tolazamide treatment the fasting plasma glucose fell from 180 +/- 19 to 134 +/- 13 mg/dL. Despite dramatic potentiation of glucose-stimulated insulin release and further improvement of early phase insulin release, K values again failed to progressively rise. This paradox occurred even in three additional subjects tested after two years of tolazamide treatment, suggesting that tolazamide may not ameliorate the postreceptor defects that impede the expression of the Staub effect. The applicability of the glucose facilitatory effect to the treatment of NIDDM might be limited to subjects in whom adjunctive measures have reestablished effective tissue responsiveness to endogenous insulin.