Percutaneous transluminal angioplasty was performed in 39 consecutive patients with atheromatous renal artery stenosis associated with hypertension. The mean blood pressure before angioplasty was 191/107 mmHg and this had dropped to a mean of 167/90 mmHg at the patient's most recent visit, representing a significant fall in both systolic (p less than 0.01) and diastolic pressures (p less than 0.001). The mean serum creatinine was 166.7 mumol/l before percutaneous transluminal angioplasty and 155.3 mumol/l at the most recent visit (not statistically significant). The mean number of anti-hypertensive drugs fell from 2.4 to 1.9 after percutaneous transluminal angioplasty (p less than 0.05). Three patients (eight per cent) were 'cured' (diastolic blood pressure less than 90 mmHg without medication), 25 (64 per cent) had 'improved' (diastolic blood pressure less than 109 mmHg, with a fall of more than 15 per cent) and 11 (28 per cent) had not improved. Logistic discriminant analysis showed that pre-percutaneous transluminal angioplasty diastolic blood pressure, age, serum creatinine and smoking habit together correctly predicted the outcome of percutaneous transluminal angioplasty in 90 per cent of patients, with four 'false positives' and no 'false negatives'. Ten patients suffered a total of 12 serious complications related to the procedure: one death in acute renal failure, one myocardial infarction, one severe hypotension just after the procedure, one deep vein thrombosis, one episode of transient ischaemia of the toes and seven groin haematomas. Thus percutaneous transluminal angioplasty for atheromatous renal artery stenosis rarely 'cures' hypertension, but improved blood pressure control is often achieved, albeit at the expense of troublesome complications. A prospective, randomized trial is needed to establish whether or not the improvement is due directly to percutaneous transluminal angioplasty.