OBJECTIVE To assess the safety and clinical and immunological activity of zalcitabine/zidovudine combination therapy compared with zidovudine monotherapy in persons with no or limited antiretroviral experience and CD4 counts of 300-500 cells/mm3. METHODS A double-blind controlled multi-centre study conducted in specialist human immunodeficiency virus (HIV) care centres in Spain, Portugal and Australia. Participants were randomized at study entry to zidovudine (200 mg three times daily) plus zalcitabine (0.75 mg three times daily) or matched placebo. The primary end point was the proportion of patients with CD4 above baseline value at 24 months. The secondary end points were time to AIDS/death, quality of life (by MOS-30) and safety. RESULTS The study was terminated prematurely following the results of the Delta and ACTG 175 studies. Two-hundred and fifty-six patients entered the protocol of whom all but 15 were treatment naive. One hundred and twenty-seven patients commenced zidovudine and 129 commenced a combination of zidovudine/zalcitabine. The median duration of follow-up was 634 days with a median time on blinded therapy of 500 days. Using the last available CD4 count data, 32.4% randomized to zidovudine and 65.1% randomized to zidovudine/zalcitabine remained above baseline at study close (P < 0.001). No significant differences were observed in the time taken for CD4 count to return to baseline. Over 104 weeks, median CD4 counts rose in the combination group from 399 to 509 cells/mm3 whereas those randomized to zidovudine fell from 410 to 374 cells/mm3. Only 12 AIDS events and two deaths (both accidental) occurred during the study with no differences between groups. No differences in quality of life were observed. Adverse events were the cause of treatment discontinuation in 8.6% of patients with no differences between treatment arms. A further 8.2% of patients were lost to follow-up. At least one adverse event (all severities, all relationships) was experienced by 80.3% of patients randomized to zidovudine and 79.8% of patients on combination. Peripheral neuropathy (all grades) was reported in 10.1% of patients randomized to the combination and 3. 1% in the zidovudine arm (P = 0.026). Oral ulcers were reported in 7.8% and 4.7% of combination and zidovudine monotherapy arms, respectively. Neutropenia was more common in the zidovudine group (22%) than the combination group (14%). CONCLUSIONS Combination of zidovudine/zalcitabine as initial therapy maintains CD4 count above commencement levels in a significantly greater proportion of patients than zidovudine monotherapy. In persons with CD4 counts > or = 300 cells/mm3 inclusion of zalcitabine with zidovudine does not increase the incidence of adverse events or adversely affect quality of life.