The usefulness of screening patients for renovascular hypertension by infusion of saralasin, a competitive antagonist of angiotensin II, was evaluated. Responses were compared in 19 patients with proved renovascular hypertension and in 34 without renovascular hypertension, as indicated by renal arteriography and renal venous renin studies. Saralasin infusion was carried out in the morning after furosemide, 80 mg by mouth, had been given the previous evening. Seventy-five percent of patients with and 12 percent of those without renovascular hypertension had a reduction in diastolic pressure of 5 mm Hg or more during saralasin infusion; only 45 percent of patients with and 6 percent of those without renovascular hypertension had a reduction of 10 mm Hg or greater during infusion. In comparison, 80 percent of patients with and 18 percent of those without renovascular hypertension had a positive intravenous pyelogram. The predictive value of a positive saralasin test (5 mm Hg or greater reduction in diastolic pressure) was calculated for varying prevalence rates of renovascular hypertension with use of Bayes theorem. The results indicate that when the prevalence rate of renovascular hypertension among hypertensive patients is 5 percent only 25 percent of positive saralasin tests will correctly predict its presence.