In insulin-dependent diabetes mellitus (IDDM), microalbuminuria predicts renal and cardiovascular disease. We report a combined analysis of 235 normotensive IDDM patients with microalbuminuria who participated in two 24-month double-blind, randomised, placebo-controlled trials to assess the effects of captopril 50 mg twice daily on the progression to overt clinical albuminuria. Of the 225 patients who were evaluable on an intent to treat basis, 25 of 114 placebo-treated patients (21.9%) and 8 or 111 captopril-treated patients (7.2%) progressed to persistent clinical albuminuria. The risk of progression over 24 months was significantly reduced by captopril (p = 0.004) with a risk reduction of 69.2% (95% confidence interval (CI):31.7 to 86.1%). This degree of risk reduction remained at the same level (62.9% [16.1-83.6%], p = 0.017) after adjustment for differences in time-varying mean arterial blood pressure. Albumin excretion rate increased by an average of 14.2% [3.1-26.5%] per year in the placebo-treated group compared with a reduction of 9.6% [-18.6-0.4%] per year in the captopril-treated group (p = 0.002). The rate of fall of creatinine clearance tended to be faster in the placebo-treated group than in the captopril-treated group (-6.4 [-10.2- -2.5] vs -1.4 [-5.3-2.6] ml.min-1.1.73m-2, p = 0.07). Baseline albumin excretion rate (p < 0.0001) and glycated haemoglobin (p = 0.03) were independent predictors of progression to clinical albuminuria and changes in mean arterial blood pressure (p = 0.02) and serum cholesterol level (p = 0.003) were significantly associated with percentage changes in albumin excretion rate. Captopril reduces the risk of progression to overt nephropathy in IDDM patients with microalbuminuria, an effect partly independent of its blood pressure-lowering effects.