Urinary C-peptide excretion has been found to be an accurate index of insulin secretion under a variety of physiologic conditions, such as acute starvation and exercise, and after oral and intravenous glucose administration. We investigated urinary C-peptide responses in a group of patients who were receiving all of their nutrient intake by intravenous administration. In these patients receiving total parenteral nutrition (TPN), we were able to monitor changes in insulin secretion when the same nutrients were infused at different rates, for example, during cyclic vs. continuous TPN administration, and to observe changes in the insulin secretory response as the pattern of nutrient delivery was altered in the same individual. We found that increasing the TPN infusion rate by 50% during cyclic TPN caused a 65% increase in serum insulin levels over levels observed during continuous TPN administration (93 vs. 60 microU/ml), whereas a 100% increase in the cyclic TPN infusion rate above the continuous TPN rate increased insulin levels by 147% (147 vs. 60 microU/ml). The molar ratio of insulin to C-peptide was increased by increasing rates of TPN infusion, from 0.116 during fasting periods to 0.151 during maximum rates of TPN administration. An additional finding of this study is that 24-hour insulin secretion, estimated by urinary C-peptide measurements, was equivalent in all treatments regardless of the pattern of insulin response elicited.