We conducted a multicenter trial of treatment with all-trans retinoic acid (ATRA) for newly diagnosed acute promyelocytic leukemia (APL) in the AML-92 study and compared it with our previous study with standard intensive chemotherapy, the AML-89 study, in the view of complete remission (CR) rate, incidence of early death, and event-free survival (EFS). Patients were scheduled to receive oral ATRA 45 mg/m2 daily until CR. If patients had leukocyte counts above 3 x 10(9)/L at the start of therapy, they received daunorubicine (DNR) 40 mg/m2 for 3 days and behenoyl cytosine arabinoside (BHAC) 200 mg/m2 for 5 days in addition to ATRA. During the ATRA therapy, if patients showed myeloblast plus promyelocyte counts higher than 1 x 10(9)/L in the peripheral blood, they received additional DNR and BHAC in the same schedule, as well. A total of 110 patients were entered into the study. Median age was 43 years (range, 16 to 74). Twenty-eight (26%) of 109 patients (one died before the start of therapy) received ATRA alone. Ninety-seven patients (89%) achieved CR; 48 of 49 (98%) aged less than 40 years, 44 of 52 (84%) aged between 40 and 69, and 5 of 8 (63%) aged above 70 achieved CR, respectively; 25 of 28 (89%) with ATRA alone, 46 of 51 (90%) with ATRA plus initial chemotherapy and 26 of 30 (87%) with ATRA plus later chemotherapy attained CR, respectively. Nine (8%) patients died within 28 days after the start of therapy. In contrast, 44 of 62 patients (71%) attained CR, and 13 (21%) died within 28 days in the AML-89 study with the combination of DNR, BHAC, 6-mercaptopurine and prednisolone. Seven developed retinoic acid syndrome and one died of it in the present study. Other toxicities associated with this drug included cheilitis, desquamation, muscle pain, and hypertriglyceridemia. Predicted 23 months EFS for all ATRA-treated patients and disease-free survival (DFS) in the CR cases were 75% and 81%, respectively, in a median follow-up period of 21 months. Compared to the AML-89 study, there was a highly significant difference in remission rate (P = .004), EFS (P = .0007), and also early mortality rate (P = .02). Present results demonstrated that ATRA with or without chemotherapy gives a statistical improvement in CR rate and early mortality rate, as well as superior survival in newly diagnosed APL.