To improve understanding of the relationships between gastric inhibitory polypeptide (GIP) and insulin secretion and food intake in obesity, immunoreactive insulin and immunoreactive GIP were measured in 5 obese children during PO glucose tolerance test carried out before and after diet. Before diet, mean insulin levels were normal at fasting and rose after glucose ingestion. The mean fasting immunoreactive GIP level was very high (1235 +- 209 pg/ml) compared to that of 8 healthy adult controls (411 +/- 44 pg/ml) and remained at this level throughout the test. There was only a short postabsorptive rise to 1515 +/- 158 pg/ml at 30 min, which was not significantly different either from the patients' basal values or from the 30-min control values (1356 +/- 67 pg/ml). After dieting for 3 to 7 months, immunoreactive insulin responses returned to normal ranges. Concomitantly, both basal and total GIP release diminished significantly (basal GIP, 343 +/- 92 pg/ml; area under the GIP curve, 3820 and 1694 pg/ml/hr before and after diet, respectively). The postabsorptive GIP increment, however, rose significantly from 180 pg/ml/hr, before diet, to 665 pg/ml/hr afterwards. These results might be compatible with the hypothesis that in obesity, hyperinsulinemia, and overactivity of the GIP cells are associated phenomena caused by overeating and reversed by reduced food intake. However, several contradictory findings remain unexplained. The discrepancy between insignificant postabsorptive GIP increments and elevated insulin responses before diet casts doubts on the causal relationship between GIP and insulin secretion. The small GIP rise might be due to a limited secretory capacity of the GIP cells or to a diminished stimulatory capacity of glucose. The constantly high level of GIP might reflect chronic hypersecretion and/or some defect in basal regulation and feedback control of GIP release. The change caused by dietary measures in the GIP secretion pattern provides evidence that in obese children, basal GIP secretion in influenced by nutritional factors.