OBJECTIVE To study the effect of acarbose, an alpha-glucosidase inhibitor, on postprandial plasma glucose and insulin and insulin sensitivity in subjects with impaired glucose tolerance (IGT). METHODS Subjects with IGT were randomly treated in a double-blind fashion with placebo (n = 10) or acarbose (n = 8) at 100 mg t.i.d. for 4 months. All subjects were submitted before randomization and at the end of the study to a standardized breakfast and a 12-h daytime plasma glucose and plasma insulin profile, and insulin sensitivity was measured as steady-state plasma glucose (SSPG) using the insulin suppression test. RESULTS While placebo had no effect on postprandial plasma glucose and plasma insulin incremental area under the curve (AUC) (3.03 +/- 0.5 vs. 3.76 +/- 0.6 mmol.h-1.l-1, P = NS; 1,488 +/- 229 vs. 1,609 +/- 253 pmol.h-1.l-1, P = NS), acarbose resulted in a significant reduction for both glucose (1.44 +/- 0.3 vs. 4.45 +/- 0.9 mmol.h-1.l-1, P = 0.002) and insulin (626.7 +/- 104.3 vs. 1,338.3 +/- 220.5 pmol.h-1.l-1, P = 0.003). The reduction in 12-h plasma glucose and insulin AUC on acarbose (11.2 +/- 2.1 mmol.h-1.l-1 and 7.5 +/- 0.7 nmol.h-1.l-1) was significantly greater than that on placebo (4.0 +/- 1.6 mmol.h-1.l-1 and 0.8 +/- 0.4 nmol.h-1.l-1) (P = 0.014 and 0.041). While SSPG was not affected by placebo (13.9 +/- 0.4 vs. 13.8 +/- 0.3 mmol/l; P = NS), it was significantly improved by acarbose (10.9 +/- 1.4 vs. 13.1 +/- 1.5 mmol/l, P < 0.004) and was also significantly different from placebo at 4 months (P < 0.02). CONCLUSIONS It is concluded that in subjects with IGT, acarbose treatment decreases postprandial plasma glucose and insulin and improves insulin sensitivity. Acarbose may therefore be potentially useful to prevent the progression of IGT to NIDDM.