Screening for renal artery stenoses in hypertensive patients aims at detecting lesions whose treatment (renal revascularization) will normalize or reduce blood pressure and correct or prevent reduced glomerular filtration. Consequently, screening tests such as renal artery duplex Doppler scanning, renal scintigraphy or digital-subtraction angiography are used in patients in whom hypertension is severe, drug-resistant or associated with renal failure. Surgical repair or transluminal angioplasty is not warranted for all stenoses, however, particularly in atheromatous stenoses where these procedures have a 1% mortality, a 10% morbidity and a 30% failure rate to improve blood pressure despite adequate anatomical outcome. Predictors of favourable blood pressure outcome following revascularization are aetiological (fibrous dysplasia rather than atheroma), historical (young age, short duration of hypertension), physiological (renal ischaemia confirmed by scintigraphy, lateralizing renal vein renin ratio) and anatomical (truncal rather than ostial or branch stenoses). Outcome of surgery and transluminal angioplasty has only been documented in retrospective, uncontrolled reports in which blood pressure improvement is overestimated via the placebo effect, habituation to blood pressure readings and optimization of drug treatment, the latter being frequently required despite adequate revascularization. The first prospective randomized trials evaluating angioplasty in atheromatous stenoses are underway and should provide objective information concerning the risk/benefit ratio of this procedure.