There is controversy about whether chemotherapy or an HLA-identical sibling bone marrow transplant is better treatment for adults with acute lymphoblastic leukemia (ALL) in first remission. A previous study of patients treated in 1980-1987 showed similar leukemia-free survivals with these approaches. We re-examined this issue in more recently treated patients receiving different chemotherapy. Chemotherapy subjects (n = 76) participated in trial ALL-87 of the Japan Adult Leukemia Study Group (JALSG). Transplant subjects (n = 214) were reported to the International Bone Marrow Transplant Registry (IBMTR). Treatment-related mortality, relapse and leukemia-free survival were compared after adjusting for differences in subject- and disease-related variables and time-to-treatment. Outcomes differed in persons < or = and >30 years of age. Five-year treatment-related mortality in persons < or =30 years was 3% (95% confidence interval, 0-12%) with chemotherapy vs 32% (23-41%; P < 0.0001) with transplants. The difference was greater among persons >30 years, 13% (2-31%) with chemotherapy vs 57% (43-69%; P < 0.0001) with transplants. Five-year relapse probability in persons < or =30 years was 69% (50-84%) with chemotherapy vs 22% (14-32%; P < 0.0001) with transplants. Among persons >30 years, 5-year relapse was 70% (53-85%) with chemotherapy vs 32% (20-45%; P < 0.0001) with transplants. Leukemia-free survival at 5 years was significantly worse with chemotherapy than with transplants in persons < or =30 years (30% (15-48%) vs 53% (44-63%; P = 0.02)) but not in persons >30 years (26% (13-41%) vs 30% (20-41%; P = 0.70)). We concluded that transplants result in more treatment-related deaths but fewer relapses than chemotherapy. Leukemia-free survival is better with transplants than chemotherapy in persons < or =30 years of age but comparable in older persons.