The criteria for selection of patients who may benefit from reconstruction of totally occluded renal arteries are not well defined. Of those patients who underwent renal artery reconstruction for renal artery occlusive disease at our institution 13% (52/397) had a chronic renal artery occlusion at preoperative arteriography. In 38 cases (73%) the contralateral renal artery had a significant (greater than 70%) stenosis, which was reconstructed simultaneously. Renovascular hypertension alone (n = 27;52%) or in combination with impaired renal function (n = 25;48%) was the indication for surgical treatment and transaortic thromboendarterectomy was the preferred method of reconstruction (n = 38;73%). Operative mortality was 5.7% and during follow-up (38 +/- 31 months) 4 patients died from myocardial infarction (n = 2), aortic dissection (n = 1) or cerebral haemorrhage (n = 1). Postoperatively 20 patients (44.5%) had a normal blood pressure; in a further 20 (44.5%) the hypertension was easier to control and medication could be reduced. Five patients (11%) remained hypertensive. Postoperative arteriography showed a normal patent renal artery in 40 cases (89%). Postoperative plasma creatinine levels ranged from 0.8 to 3.4 mg% (1.32 +/- 0.74) which was a significant reduction from preoperative levels. Two patients with severe uraemia improved to an extent where haemodialysis could be discontinued. Radionuclide scan and the measurements of plasma renin activity before and after administration of a converting enzyme inhibitor (Captopril) were the most reliable diagnostic methods for preoperative patient selection and postoperative follow-up.(ABSTRACT TRUNCATED AT 250 WORDS)