Carbohydrate intolerance is positively correlated with animal fat consumption and is more common in beef eating populations. In contrast, individuals consuming diets comprised of polyunsaturated fats have a lower incidence of diabetes mellitus. To test the hypothesis that dietary fats may influence carbohydrate metabolism, serum glucose, insulin, and gastric inhibitory polypeptide (GIP) responses to three mixed test meals of varying fatty acid composition were assessed in 12 normal subjects. Fatty acids in the meals were either saturated fats or polyunsaturated fats derived from vegetables or fish. Each test meal provided 40% of a subject's calculated daily caloric requirement and contained approximately 45% carbohydrate, 40% fat, and 15% protein. Serum insulin responses were 62% higher (p less than 0.01) after the fish and 39% higher (p less than 0.01) after the vegetable meals compared to the saturated fat meal. No significant differences in insulin responses were observed between the vegetable and fish meals. Serum glucose concentration was slightly higher (p less than 0.02) during the fish meal than with the vegetable or saturated fat meals. The GIP levels were comparable following the fish and vegetable meals and were 25% lower than those observed with the saturated fat meal. These findings suggest that diets enriched with polyunsaturated fatty acids augment insulin secretion significantly more than a diet comprised primarily of saturated fatty acids. The mechanism for this increased insulin secretion is unknown but did not appear to be mediated through differences in serum glucose values or through the insulin-otrophic effects of GIP.(ABSTRACT TRUNCATED AT 250 WORDS)