We have evaluated the feasibility of monitoring the 24-hour urinary excretion rate of C-peptide (U-CPR) as a measure of integrated beta-cell function in patients with non-insulin-dependent diabetes mellitus (NIDDM). In 37 normoalbuminuric patients, U-CPR of 117.9 +/- 9.1 micrograms/d (mean +/- SEM) during the poorly controlled glycemic phase (fasting plasma glucose [FPG], 171 +/- 7 mg/dL; hemoglobin A1C [HbA1c], 8.8% +/- 0.4%) was significantly higher than the value of 83.3 +/- 13.7 micrograms/d (P < .001) during the well-controlled phase (FPG, 135 +/- 6 mg/dL; HbA1c, 7.0% +/- 0.2%), although the plasma insulin response to meals was lower during the former phase (53.3 +/- 6.3 microU/mL) versus the latter phase (65.7 +/- 6.6, P < .005). Endogenous creatinine clearance (Ccr) was significantly elevated during the poorly controlled phase (105.4 +/- 7.3 v 88.7 +/- 4.7 mL/min, P < .005). In 26 microalbuminuric patients, the plasma insulin response was greater during good glycemic control, but U-CPR did not differ between the two phases. Ccr was comparable at two phases in this group (92.7 +/- 7.4 v 91.1 +/- 5.9 mL/min, NS). U-CPR correlated positively with Ccr in both groups (r = .593, P < .001 in normoalbuminuria; r = .585, P < .001 in microalbuminuria). In addition, when biosynthetic human C-peptide was infused intravenously at an identical rate in two healthy subjects, resulting steady-state plasma levels of CPR were lower, and fractional U-CPR was higher during the moderately hyperglycemic phase versus the euglycemic phase.(ABSTRACT TRUNCATED AT 250 WORDS)