To examine the clinical role of BAYm 1099, 15 diet-treated non-insulin-dependent diabetic (NIDDM) subjects were randomized to start drug (50 mg 3 times/day) or placebo after a 4-wk run-in period in a double-blind crossover study. Treatment periods (4 wk) were separated by a 2-wk washout period. During the last week of each treatment period, three test meals (TMs) were administered: 60 g starch (TM1), 25 g sucrose (TM2), and combined 60 g starch and 25 g sucrose (TM3). Twelve subjects completed the study. The peak postprandial blood glucose, lactate, and pyruvate levels (means +/- SE) were significantly lower with active drug after all test meals, particularly TM2 (11.3 +/- 1.0 vs. 14.3 +/- 1.4 mM, P less than .001; 1.53 +/- 0.20 vs. 2.48 +/- 0.17 mM, P less than .001; and 105.1 +/- 17.6 vs. 147.6 +/- 11.1 microM, P less than) less than .001; and 105.1 +/- 17.6 vs. 147.6 +/- 11.1 microM, P less than .05, respectively. Peak blood glucose levels were significantly delayed. However, fasting blood glucose, HbA1, fructosamine, and cholesterol did not change during active treatment (10.0 +/- 1.0 vs. 9.9 +/- 1.0 mM, 10.0 +/- 0.7 vs. 9.4 +/- 0.7%, 2.44 +/- 0.10 vs. 2.37 +/- 0.07 mmol/100 g protein, and 6.7 +/- 0.3 vs. 6.5 +/- 0.3 mM, P NS). Flatulence and diarrhea were severe in 2 subjects, requiring termination of study. Thus, in NIDDM, BAYm 1099 was effective in diminishing and delaying postprandial excursions of blood glucose, lactate, and pyruvate after high- and low-sucrose meals, but overall metabolic control remained unchanged.(ABSTRACT TRUNCATED AT 250 WORDS)