To determine if source and amount of carbohydrate affected postprandial glucose and insulin responses, seven nondiabetic subjects consumed 0, 25, 50, 75 or 100 g carbohydrate (total carbohydrate minus total dietary fiber) portions of barley, spaghetti, bread or potato. By ANOVA, both source and amount of carbohydrate had significant effects on incremental response areas for capillary glucose (P = 0.001), plasma glucose (P = 0.01) and plasma insulin (P = 0.03), but there was no source x amount interaction. By regression analysis, source of carbohydrate explained a similar amount of the variability of glucose and insulin responses, 46-64%, as the amount of carbohydrate, 47-57%. Together, carbohydrate source and amount accounted for 85-94% of the variability of mean glucose and insulin responses. We conclude that, for individual foods with different glycemic indices, both source and amount of carbohydrate influence the postprandial glucose and insulin responses of nondiabetic subjects.