OBJECTIVE To determine the efficacy of acarbose, compared with placebo, on the metabolic control of NIDDM patients inadequately controlled on maximal doses of conventional oral agents. RESEARCH DESIGN AND METHODS In this three-center double-blind study, 90 Chinese NIDDM patients with persistent poor glycemic control despite maximal doses of sulfonylurea and metformin were randomly assigned to receive additional treatment with acarbose 100 mg thrice daily or placebo for 24 weeks, after 6 weeks of dietary reinforcement. Efficacy was assessed by changes in HbA1c, fasting and 1-h postprandial plasma glucose and insulin levels, and fasting lipid levels. RESULTS Acarbose treatment was associated with significantly greater reductions in HbA1c (−0.5 ± 0.2% vs. placebo 0.1 ± 0.2% [means ± SEM], P = 0.038), 1-h postprandial glucose (−2.3 ± 0.4 mmol/1 vs. placebo 0.7 ± 0.4 mmol/1, P < 0.001) and body weight (−0.54 ± 0.32 kg vs. placebo 0.42 ± 0.29 kg, P < 0.05). There was no significant difference between the two groups regarding changes in fasting plasma glucose and lipids or fasting and postprandial insulin levels. Flatulence was the most common side effect (acarbose vs. placebo: 28/45 vs. 11/44, P < 0.05). One patient on acarbose had asymptomatic elevations in serum transaminases that normalized in 4 weeks after acarbose withdrawal. Another patient on acarbose developed severe hypoglycemia; glycemic control was subsequently maintained on half the baseline dosage of sulfonylurea. CONCLUSIONS In NIDDM patients inadequately controlled on conventional oral agents, acarbose in moderate doses resulted in beneficial effects on glycemic control, especially postprandial glycemia, and mean body weight. Additional use of acarbose can be considered as a useful alternative in such patients if they are reluctant to accept insulin therapy.