ObjectiveTo evaluate the efficacy and safety of zidovudine (ZDV) at a maintenance dose of 250 mg every 6 h alone or as cotherapy with acyclovir (ACV; 800 mg every 6 h) as treatment for AIDS and AIDS-related complex (ARC).DesignDouble-blind, randomized, placebo-controlled clinical trial of up to 1 year's therapy.SettingTeaching hospital ambulatory clinics in eight European countries and Australia.SubjectsA total of 131 patients with AIDS and 134 with ARC were enrolled and followed from 1986 to 1988.Main outcome measuresTime to development of AIDS-defining opportunistic infections and AIDS-associated neoplasms, survival assessed at 1 year after entry, performance status, body weight, CD4 + cell counts.ResultsDuring the study period, 46 (36%) ZDV recipients and 37 (27%) cotherapy recipients developed opportunistic infections. The probability of an ARC patient progressing to AIDS (1982 Centers for Disease Control criteria) was 0.18 and 0.15 [95% confidence interval (CI) for difference, —0.17 to 0.11] for the ZDV alone and cotherapy recipients, respectively. After excluding patients who experienced an opportunistic infection during the first 4 weeks of therapy, the probability was 0.13 and 0.099 (95% Cl for difference, —0.16 to 0.10) for the ZDV and cotherapy recipients, respectively. Thirty-six patients treated with single-agent therapy [28 (41%) AIDS and eight (12%) ARC patients] and 15 cotherapy recipients [13 (21%) AIDS and two (3%) ARC patients] died during the study. There was a significant difference in time to death between the cotherapy and ZDV alone groups for both AIDS (P = 0.014) and ARC (P = 0.045) patients, with cotherapy patients surviving longer. Infections related to herpesviruses, but not cytomegalovirus, were reduced in patients receiving ACV therapy. CD4 + cell counts in both arms generally increased initially and then declined. Forty-six per cent of patients in the ZDV group (59% of AIDS and 31% of ARC patients) and 52% of patients in the cotherapy group (69% of AIDS and 34% of ARC patients) experienced bone-marrow suppression. Red cell transfusions were administered to 33% of ZDV alone recipients and 34% of cotherapy recipients.ConclusionThese data show that the addition of high-dose ACV cotherapy to ZDV for patients with AIDS and advanced ARC results in a statistically significant improvement in survival with minimal increase in the risk of toxicity.