With its incidence of about 40%, acute renal failure (ARF) is a major problem after cadaveric renal transplantation. We have previously shown that, with moderate hydration (2.5 L) of the recipient, together with rapid infusion of 250 ml of mannitol 20% just before clamp removal, the incidence of ARF decreased to below 10%. Administration of mannitol without hydration was not effective. In a prospective randomized trial we have now investigated whether hydration without mannitol is sufficient to prevent ARF. For this purpose patients were randomly allocated to treatment with moderate hydration with or without mannitol. Furthermore, in both treatment groups recipients were randomized to treatment with cyclosporine or azathioprine. The allocation method used guaranteed an even distribution for 10 important prognostic factors. In the cyclosporine group, the percentage of ARF was significantly lower in mannitol-treated (n = 32) than in glucose-treated patients (n = 32) (19% vs. 54%, P less than 0.01). In the azathioprine group the percentage of ARF was also lower in mannitol-treated (n = 33) than in glucose-treated patients (n = 34) (18% vs. 44%, P less than 0.05). Overall incidence of ARF in both groups was significantly lower in mannitol-treated patients (P less than 0.001). Thus, moderate hydration and administration of 250 ml mannitol 20% just before arterial clamp removal are both indispensable for optimal prevention of ARF after cadaveric renal transplantation.