The mechanism of action of angiotensin-converting enzyme (ACE) inhibitors on urinary albumin excretion (UAE) in diabetic patients remains controversial. Sixteen type 1, insulin-dependent diabetics with incipient nephropathy received ramipril, a long-acting ACE inhibitor, at hypotensive doses (treatment A: 5 mg/day, n = 8) or at nonhypotensive doses (treatment B: 1.25 mg/day, n = 8) during a 6-week, double-blind, parallel study to establish whether its antihypertensive effects could be dissociated from its local renal effects. Blood pressure, UAE, glomerular filtration rate (GFR), effective renal plasma flow (ERPF, constant [I-125]iodothalamate + [I-131]hippurate infusion), and ACE activity were measured before and after treatment. Blood pressure was lowered with treatment A but not with treatment B. UAE and ACE activity were reduced with both treatments. Baseline GFR and ERPF were not altered by either treatment. In the patient population as a whole, ACE inhibition correlated with a rise in ERPF and with a reduction in filtration fraction (GFR/ERPF), but not with the changes in blood pressure. Changes in UAE correlated with the changes in filtration fraction. It is concluded that renal hemodynamics may be modified by ramipril independently of blood pressure changes.