Proinsulin is converted to insulin and C-peptide in the pancreatic beta-cells; the latter two peptides are secreted in equimolar concentrations. Thus measurements of C-peptide immunoreactivity may provide a means of assessing residual pancreatic function in insulin-treated diabetic patients. Thirty-five patients with a mean (+/- SE) age of 13.4 +/- .6 years who had diabetes mellitus for 4.8 +/- .3 years were included in this study. Glucose and CPR were measured in the fasting state and one hour after 1 gm/kg (maximum 50 gm) of oral and glucose. Patients were assigned to one of two groups on the basis of adequate or poor control of diabetes. Twenty-five of the 35 (71%) patients had evidence of endogenous beta-cell function, i.e., CPR greater than 0.5 ng/ml. CPR levels over 0.5 ng/ml were present in a significantly (p less than 0.05) greater number of patients with diabetes of less than 5 years duration (19/21) than in those with diabetes greater than 5 years duration (6/14). Only one patient showed a rise in CPR after the glucose load. All patients with CPR greater than 2.0 ng/ml were in the adequately controlled groups, but there were patients with CPR less than 2.0 ng/ml in both adequately and poorly controlled groups. Because the CPR value includes both C-peptide and antibody-bound proinsulin, separate determination of free C-peptide was done in 30 patients. These results confirmed the conclusions based on CPR estimation. Although growth hormone values were higher in patients in the poorly controlled group, there was no correlation between hGH and CPR. We conclude that residual insulin secretion in diabetic patients may facilitate good control, but that low CPR values and hence absent beta-cell reserve is not always associated with poor control.