Twelve normal subjects and 10 subjects with non-insulin-dependent diabetes mellitus were given, in random order at intervals of greater than or equal to 1 wk, three drinks of the same beverage: one unsweetened, one sweetened with 400 mg aspartame, and one sweetened with 135 mg saccharin. The amount of sweetener approximated that in 1 L of sugar-free soft drink. Plasma glucose, insulin, and glucagon were measured for 3 h after ingestion of the test beverage. Plasma glucose declined slightly throughout the test period, probably due to fasting, with no differences between the three treatments. Neither sweetener affected peak insulin levels in subjects with or without diabetes. Analysis of area under the curve showed that mean insulin levels were statistically significantly higher after aspartame than after saccharin or unsweetened beverage in normal subjects only, but the magnitude of the difference was small and unlikely to be of physiological importance in the absence of differences in glucose levels. Furthermore, the differences could largely be accounted for by a decrease in insulin values after both unsweetened beverage and saccharin, with no change from baseline after aspartame. Glucagon levels showed time-to-time variation but no overall differences. We conclude that ingestion of aspartame- or saccharin-sweetened beverages by fasting subjects, with or without diabetes, did not affect blood glucose homeostasis.