A prospective randomized study on aggressive non-Hodgkin's lymphomas was conducted by investigators at several Italian institutions with the intent of comparing two third-generation conceptually different regimens: the regimen containing methotrexate with leucovorin rescue, doxorubicin, cyclophosphamide, vincristine, prednisone, and bleomycin (MACOP-B), a short-term continuous twelve-week therapy, and F-MACHOP (5-fluorouracil, methotrexate with leucovorin rescue, cytarabine, cyclophosphamide, doxorubicin, vincristine, and prednisone), a monthly intensive cyclic treatment combining prednisone with six active non-cross-resistant cytotoxic drugs. The goals of this study were the response rate, relapse-free survival, and incidence of hematologic and nonhematologic toxicities. Two hundred-eighty-six patients included between 15 and 60 years fulfilled the criteria for entry to the study; 140 patients were treated with MACOP-B and 146 with F-MACHOP. The minimum follow-up was 24 months. Clinical characteristics of all patients were similar and known prognostic factors were equally distributed between the two groups. Complete remission (CR) was achieved by 61% and 67% of the patients treated with MACOP-B and F-MACHOP, respectively; 4% and 6% were primarily resistant, 2% and 5%, respectively, died of causes directly related to therapy. At 50 months, 74% of all CR patients were alive without disease and there were no significant differences in relapse-free survival between the two groups: 75% in the F-MACHOP group and 73% in the MACOP-B group at 50 months. There was a higher incidence of mucositis among patients treated with MACHOP-B than among those given F-MACHOP (11% vs 3.5%).(ABSTRACT TRUNCATED AT 250 WORDS)