The efficacy and tolerability of acarbose were examined in a postmarketing surveillance study of 10,462 patients (829 insulin-dependent diabetes mellitus (IDDM), 9,440 non-insulin-dependent diabetes mellitus (NIDDM), 193 not classified) during a 12-week treatment period. The median duration of diabetes was 60 months for men and 72 months for women in IDDM patients, and 40 months for men and 60 months for women in NIDDM patients. Of the Type II patients, 28.9% were treated with diet only; 58.1% additionally with sulfonylureas; 8.6% with insulin; and 4.3% with both sulfonylureas and insulin. The additional acarbose therapy led to a reduction of the mean fasting blood glucose levels (51 mg/dL for IDDM; 52 mg/dL for NIDDM) and 1 h postprandially (55 mg/dL for IDDM; 63 mg/dL for NIDDM). The HbA1 levels were reduced by 1.5%. Tolerability was good: 78.6% of patients had no adverse events; 19% reported meteorism/flatulence; 3.2%, diarrhea. Hypoglycemia was found in 0.8% of Type I and 0.6% of Type II patients who received concurrent insulin (n = 8) or glibenclamide (n = 1) treatment. Laboratory investigations gave no indication of other adverse effects, e.g. elevated levels of transaminases or creatinine. This postmarketing surveillance study documents the therapeutic benefit and the good tolerability of acarbose.