Fifty patients with congestive heart failure received, by infusion, 15 ml/kg body weight water load, and systemic hemodynamic, renal function, and neurohumoral parameters values were measured before, 2 days, and 1 month after randomly allocating patients to prazosin or captopril therapy. Both prazosin and captopril caused similar and persistent hemodynamic changes, but important differences existed between their renal and neurohumoral effects. After 1 month of continuous therapy, captopril increased creatinine clearance from 71 to 84 ml/min/1.732(p < .05), increased the water load excreted in 5 hr from 50% to 71% (p < .005), and increased 5 hr sodium excreted from 6.8 to 14.7 meq (p < .005). Captopril also caused a decrease in plasma norepinephrine from 568 to 448 pg/ml (p < .005), in plasma epinephrine from 94 to 73 pg/ml (p < .05), and in plasma aldosterone from 57 to 28 ng/dl (p < .005), without changing plasma vasopressin. These beneficial effects were greater after 1 month of therapy than after 2 days. The only beneficial effect of prazosin was to increase water excretion from 49% to 59% (p < .05). The long-term response to captopril was similar in patients with higher (> 2.5 ng/ml/hr) and lower renin levels. However, in patients with lower renin levels, prazosin decreased pulmonary capillary wedge pressure (24.8 to 21.8 mm Hg, p < .05), decreased plasma arginine vasopressin (1.16 to 0.75 pg/ml, p < .05), increased water excretion (62% to 85%, p < .005), and decreased plasma epinephrine (81 to 46 pg/ml, p < .05), while in patients with higher renin levels none of these beneficial effects were noted. We conclude (1) that captopril produces long-term beneficial renal and neurohumoral effects that prazosin does not despite similar hemodynamic changes with the two drugs, (2) that these effects are at least partially dependent on the initial neurohumoral and hemodynamic status of the patient, and (3) that through hemodynamic improvement vasodilators may chronically interrupt vasopressin overstimulation.