A double-blind, randomised, parallel group, multicentre, multinational study compared the effects of 12 months' treatment with lisinopril (10-20 mg once daily) or nifedipine retard tablets (20-40 mg twice daily) in 239 males (aged 18-75 years) and 96 post-menopausal females (aged 40-75 years). They all had a history of clinically stable type II diabetes > 3 months, microalbuminuria and early diabetic nephropathy (a urinary albumin excretion (UAE) rate ranging from 20 to 300 micrograms/min) and a sitting diastolic blood pressure (DBP) 90-100 mm Hg (Korotkoff phase V) inclusive at both entry and after 3-4 weeks' placebo treatment. The aim of treatment was to achieve a reduction in sitting DBP to < 90 mm Hg 24-30 h after the last dose of lisinopril or 12-18 hours after the last dose of nifedipine and to evaluate the effect of these treatments on UAE over 12 months. The effect of the two treatments on ambulatory blood pressure (BP) was also evaluated in a subset of patients. Management of diabetes with oral hypoglycaemic drugs, diet and insulin alone or in combination was permitted. Median UAE fell on lisinopril from 65.5 (range 20-297) micrograms/min at baseline to 39.0 (2-510) micrograms/min after 12 months. On nifedipine median UAE fell from 63.0 (range 20-289) micrograms/min at baseline to 58.0 (9-1192) micrograms/min after 12 months. The estimated median difference between the effects of the two treatments was 20 micrograms/min (P = 0.0006). Over 12 months both treatments produced similar falls in sitting BP from 163 +/- 17/99 +/- 6 mm Hg (mean +/- s.d.) to 147 +/- 18/88 +/- 10 mm Hg for lisinopril and from 161 +/- 18/97 +/- 5 mm Hg to 150 +/- 18/88 +/- 9 mm Hg for nifedipine. Ambulatory BP was assessed in a subset of patients and using areas under the BP-time curve (AUC) a comparison of the effects of the two treatments showed no between-treatment differences. Creatinine clearance, glycaemic control (HbA1c) and lipid profiles did not change significantly during either treatment. Frequency of withdrawals and adverse events were similar for both treatments. We conclude that lisinopril has a significantly more beneficial effect on UAE than nifedipine despite similar effects on both BP and glycaemic control in type II diabetic patients with hypertension.