Serum insulin, proinsulin and C-peptide were measured in 6 patients with chronic renal failure and 12 controls in the fasting state and following oral glucose and i.v. arginine. Fasting levels of insulin (0.13 .+-. 0.05 nM vs. 0.05 .+-. 0.01 nM), proinsulin (0.12 .+-. 0.06 nM vs. 0.02 .+-. 0.01 nM) and C-peptide (3.35 .+-. 1.35 nM vs. 0.65 .+-. 0.26 nM) were higher in patients with chronic renal failure than in controls, the increase in proinsulin (7.5-fold) and C-peptide (4.9-fold) exceeding that of insulin (2.4-fold). Following glucose, insulin rose 11.7 times (60 min) in controls, while the increment in chronic renal failure patients was less at 7.5-fold (90 min). C-peptide peaked at 3.7-fold (90 min) in controls and fell toward basal levels at 240 min. In contrast, in chronic renal failure patients, C-peptide peaked later (180 min) reaching levels of 7.51 nM (1.9-fold increase), and at 240 min remained near maximum levels (7.18 nM). The glucose stimulated C-peptide levels in patients with chronic renal failure were significantly higher (P < 0.001) than in controls at all times. With arginine stimulation in chronic renal failure patients, C-peptide and insulin increments were much lower, with C-peptide levels again being significantly higher (P < 0.01) than in controls throughout the test. C-peptide and insulin were highly correlated in the two tests in controls (r = 0.77 with glucose; r = 0.76 with arginine). In chronic renal failure, the correlation approached significance (0.05 < P < 0.1) following arginine administration, but did not correlate after glucose due to different temporal responses of the 2 peptides. Renal dysfunction causes marked changes in serum .beta. cell peptide levels, and their relative concentrations may vary depending upon the quantitative role which the kidneys play in their removal. Changes in the relationship of C-peptide to insulin must be considered when interpreting .beta. cell function in diabetics and non diabetics with chronic renal failure.