Long-term outcome was studied in 233 patients who had undergone renal artery revascularization (51 with balloon angioplasty, 182 with surgery) between 1976 and 1992. Patients (excluding renal transplants) were treated for renal vascular hypertension without or with renal insufficiency (serum creatinine > 1.6 mg/dl. All patients still alive (n = 188) were contacted to determine current blood pressure, medications, serum creatinine, and subsequent significant medical events. In patients who had died the cause of death was determined and renal function status at the time of death noted from medical records. Some follow-up information was obtained on all 233 patients; follow-up serum creatinine data were obtained in 193 (82.8%) patients. Some 24 patients (10.3%) became dialysis-dependent. Using a multiple logistic regression analysis only, preoperative creatinine maintained significance (P < 0.001) for increased dialysis risk. There was no statistically significant association of dialysis for type of revascularization (percutaneous transluminal angioplasty, autogenous artery, saphenous vein, endarterectomy or synthetic material), simultaneous or previous aortic or other vascular surgery (carotid endarterectomy, femoropopliteal bypass, etc.), pathology (atherosclerosis or fibromuscular dysplasia), number of renal arteries stenosed or treated, length of follow-up, age, coronary artery disease, congestive heart failure, stroke, chronic lung disease or type II diabetes. It is concluded that, in patients with renal artery stenosis, the timing of renal artery revascularization relative to the level of renal function is the most important determinant for long-term renal salvage.