To assess the diagnostic and prognostic value of renal venous renin levels in renovascular hypertension, 95 patients with severe unilateral renal artery stenosis were studied. Surgery (n = 52) or percutaneous transluminal dilation (n = 43) were done irrespective of renal venous renin levels. Lateralization of renin secretion as assessed by the PRA ratio and the renin secretion index was found in the majority of patients (66% and 88%, respectively). Patients with fibromuscular hyperplasia had more frequently PRA ratios < 1.5 than those with arteriosclerotic stenosis (p < 0.05). The renin secretion index proved to have a higher sensitivity (92%) and predictive value (92%) for a successful outcome of both surgery and PTA than the PRA ratio (69% and 89%, respectively), while the specificity was the same with both parameters (42% and 43%, respectively). The contralateral suppression index was most specific in predicting an unfavorable outcome. However, with all ratios used, a considerable number of false-negative and false-positive tests were observed both with surgery and PTA, a finding limiting the value of the method in selecting patients for these interventions. Other factors, such as age of the patient, kidney function and the underlying arterial disease turned out to be equally important prognostic factors. Thus, although cure after both surgery and PTA is more likely in the presence of lateralized renin secretion and contralateral suppression, the method does not allow to exclude patients with severe renal artery stenosis, hypertension and negative renal venous renin tests from these interventions. Therapeutic decisions in renovascular hypertension therefore should not depend on a single test, but rather on various clinical, radiological and laboratory findings together.