Five healthy volunteers and 6 diabetics were given a mixed test meal on two occasions--once with and once without 10 g guar flour. Addition of guar caused a 47% decrease in maximum post-prandial GIP levels, a 48% decrease in blood glucose and a 48% decrease in plasma insulin in normal subjects. In diabetics, addition of guar caused a 30% reduction in maximum post-prandial GIP and 58% decrease in blood glucose. Four normal and 6 diabetic subjects were given a predominantly carbohydrate meal, again with and without 10 g guar. Addition of guar caused a 78% decrease in blood glucose and a 59% decrease in plasma insulin in normal subjects. In diabetics addition of guar caused a 71% decrease in maximum post-prandial plasma GIP and a 68% decrease in blood glucose. Lowering of post-prandial blood glucose, plasma insulin and GIP levels by guar was statistically significant in every case. Addition of guar to the predominantly carbohydrate meal caused a decrease in total plasma GLI in both normal and diabetic subjects but reached statistical significance only in the normal subjects. There was a highly significant correlation (r = 0.83; p less than 0.0005) between peak post-prandial insulin levels in normal subjects and the corresponding plasma GIP concentration. The reduction of GIP or GLI secretion may, therefore, be partly responsible for the smaller rise in plasma insulin observed in normal volunteers when guar is added to meals.