To study the effect of angiotensin converting enzyme inhibition on the rate of progression to clinical proteinuria and the rate of change of albumin excretion rates in patients with insulin-dependent diabetes mellitus and persistent microalbuminuria. Randomized, double-blind, placebo-controlled clinical trial of 2 years' duration at 12 hospital-based diabetes centers. Ninety-two patients with insulin-dependent diabetes mellitus and persistent microalbuminuria but no hypertension. The patients were randomly allocated in blocks of two to receive either captopril, 50 mg, or placebo twice per day. Albumin excretion rate, blood pressure, glycosylated hemoglobin level, and fructosamine level every 3 months; urinary urea nitrogen excretion every 6 months; and glomerular filtration rate every 12 months. Twelve patients receiving placebo and four receiving captopril progressed to clinical proteinuria, defined as an albumin excretion rate persistently greater than 200 micrograms/min and at least a 30% increase from baseline (P = .05). The probability of progression to clinical proteinuria was significantly reduced by captopril therapy (P = .03 by log-rank test). Albumin excretion rate rose from a geometric mean (95% confidence interval) of 52 (39 to 68) to 76 (47 to 122) micrograms/min in the placebo group but fell from 52 (41 to 65) to 41 (28 to 60) micrograms/min in the captopril group, a significant difference (P < .01). Mean blood pressure was similar at baseline in the two groups and remained unchanged in the placebo group but fell significantly, by 3 to 7 mm Hg, in the captopril group. Glycosylated hemoglobin levels and glomerular filtration rate remained stable in the two groups. Captopril therapy significantly impeded progression to clinical proteinuria and prevented the increase in albumin excretion rate in nonhypertensive patients with insulin-dependent diabetes mellitus and persistent microalbuminuria.