T he pattern of endogenous insulin secretion over a 24-h period, which included three mixed meals, was evaluated in 14 normal volunteers and 15 obese subjects. Insulin secretory rates were calculated from plasma C-peptide levels using individually derived C-peptide kinetic parameters and a validated open two-compartment model of peripheral C-peptide kinetics. Insulin secretion rates were consistently elevated in the obese subjects under basal conditions (11.6 +/- 1.2 vs. 5.4 +/- 0.5 nmol/h) and in the 4 h after breakfast (139 +/- 15 vs. 63 +/- 5 nmol/4 h, P less than 0.001), lunch (152 +/- 16 vs. 67 +/- 5 nmol/4 h, P less than 0.001), and dinner (145 +/- 18 vs. 65 +/- 6 nmol/4 h, P less than 0.001). In the normal subjects, basal insulin secretion represented 50 +/- 2.1% of total 24-h insulin production, insulin secretion returned to baseline between meals, and equal quantities of insulin were secreted in the 4 h after breakfast, lunch, and dinner, despite the fact that subjects consumed half the number of calories at breakfast compared to lunch and dinner. Overall glucose responses were also similar after the three meals. In contrast, the pattern of insulin secretion in obese subjects was largely normal, albeit set at a higher level. However, the insulin secretion rate after meals did not return to baseline, and the secretion rate immediately before lunch (350.5 +/- 81.9 pmol/min) and dinner (373.6 +/- 64.8 pmol/min) was considerably higher than the secretion rate immediately before breakfast (175.5 +/- 18.5 pmol/min). In these overweight subjects, the glucose response after lunch was lower than after dinner. Analysis of individual 24-h insulin secretory profiles in the normal subjects revealed that insulin secretion was pulsatile. On average 11.1 +/- 0.5 pulses were produced in each 24-h period. The most prevalent temporal distribution of postmeal secretory pulses was two pulses after breakfast and three pulses after both lunch and dinner. Insulin secretion was also pulsatile during the period without meal stimuli: 3.9 +/- 0.3 pulses occurred during the period of overnight sampling and in the 3-h period before ingestion of the breakfast meal. In the obese subjects, the number and timing of secretory pulses was similar to those of normal volunteers, although the amplitude of the pulses was significantly greater. In both groups of subjects, greater than 80% of insulin pulses were concomitant with a pulse in glucose concentration in the postmeal period. The concomitancy rate was significantly lower in the interval without the meal stimuli, averaging 47% in both groups. Thus in obesity, although hypersecretion of insulin can be documented, the temporal pattern of secretion i s largely unaltered, which suggests that the functioning beta cell mass is enhance, but normal regulatory mechanisms influencing secretion are still operative.