AbstractIn stage IV non‐Hodgkin's lymphomas, CVP (cyclophosphamide, vincristine, and prednisone) was randomly compared to ABP (adriamycin, bleomycin, and prednisone). Of 62 patients entered into the study, 57 (CVP 27, ABP 30) were considered evaluable for comparison. In patients with liver and/or marrow involvement a second biopsy was performed to define complete remission (CR). CR occurred in 48% of patients treated with CVP and in 50% of those given ABP. The median duration of CR was 10.5 and 20.5 months, respectively. The difference is not statistically significant. Also the survival of complete responders was not significantly different between the two treatment groups. After cross‐over, secondary treatment with CVP produced an overall response rate of 40% (six of 15), compared to 50% (six of 12) obtained with ABP. In the ABP group, four patients developed a reversible interstitial penumonia. In two other patients, cardiomyopathy (fatal in one) was observed. In conclusion, although complete remission was similar in both groups, cumulative toxicity occurred in few patients given ABP. However, this combination could represent in effective alternative treatment to be used either in CVP failures or in sequence with