Background: Zidovudine therapy is of benefit in the treatment of symptomatic and asymptomatic human immunodeficiency virus (HIV) infection in persons with CD4+ cell counts of less than 500 per cubic millimeter. The efficacy, safety, and duration of benefit of zidovudine in those with 500 or more CD4+ cells per cubic millimeter are uncertain. Methods: In a double-blind, placebo-controlled trial, 993 patients with asymptomatic HIV infection and CD4+ cell counts above 400 per cubic millimeter were randomly assigned to receive zidovudine (500 mg twice daily) or placebo for three years. The primary end point was progression of disease, as defined by the development of Centers for Disease Control and Prevention (CDC) group IV disease (including recurrent oral candidiasis, hairy leukoplakia, or progressive diarrhea) or two CD4+ cell counts below 350 per cubic millimeter. This outcome measure was changed from the original end point of the acquired immunodeficiency syndrome (AIDS) or advanced AIDS-related complex to reflect changes in recommendations for management. The study was terminated after the first interim analysis. Results: Disease progression was significantly less frequent in the zidovudine group (relative risk, 0.56; 95 percent confidence interval, 0.43 to 0.75; P<0.001 by the log-rank test). The probability of disease progression at two years was 0.19 with zidovudine, as compared with 0.34 with placebo (95 percent confidence interval for the difference, -0.21 to -0.08). Progression to CDC group IV disease was reduced by half in the zidovudine recipients (relative risk, 0.49; P = 0.049) and decline in CD4+ cell counts to below 350 per cubic millimeter was reduced by 40 percent (relative risk, 0.60; P<0.001). The inclusion of early HIV disease events (oral candidiasis, oral hairy leukoplakia, and herpes zoster) as end points confirmed the effects of zidovudine on the progression of clinical disease (relative risk, 0.55; 95 percent confidence interval, 0.37 to 0.84; P = 0.004). The median duration of treatment was 94 weeks. Severe hematologic or clinical side effects were rare. Conclusions: Treatment with zidovudine benefits HIV-infected persons with CD4+ cell counts above 400 per cubic millimeter. Despite the use of doses larger than those now generally prescribed, zidovudine was well tolerated for up to three years by most of our patients.