Background. After primary treatment for cryptococcal meningitis, patients with the acquired immunodeficiency syndrome (AIDS) require some form of continued suppressive therapy to prevent relapse. Methods. We conducted a multicenter, randomized trial that compared fluconazole (200 mg per day given orally) with amphotericin B (1 mg per kilogram of body weight per week given intravenously) in patients with AIDS who had completed primary therapy for cryptococcal meningitis with amphotericin B (≥15 mg per kilogram). To be eligible, patients had to have at least two negative cultures of cerebrospinal fluid immediately before randomization. The primary end point was relapse of cryptococcal disease as confirmed by biopsy or culture. Results. Of 218 patients initially enrolled, 119 were assigned to fluconazole and 99 to amphotericin B. Twenty-three patients were found not to have met the entry criteria; six other patients assigned to amphotericin B did not receive it and were lost to follow-up. Of the remaining 189 patients, after a median follow-up of 286 days 14 of 78 receiving amphotericin B (18 percent) and 2 of 111 assigned to fluconazole (2 percent) had relapses of symptomatic cryptococcal disease (P<0.001 by Fisher's exact test). There was a difference of 19 percent in the estimated probability of remaining relapse-free at one year between the fluconazole group (97 percent) and the amphotericin B group (78 percent) (95 percent confidence interval, 7 percent to 31 percent; P<0.001). Serious drugrelated toxicity was more frequent in the amphotericin B group (P = 0.02), as were bacterial infections (P = 0.004) and bacteremia (P = 0.002). Conclusions. Fluconazole taken by mouth is superior to weekly intravenous therapy with amphotericin B to prevent relapse in patients with AIDS-associated cryptococcal meningitis after primary treatment with amphotericin B. (N Engl J Med 1992;326:793–8.). © 1992, Massachusetts Medical Society. All rights reserved.