PURPOSE To determine the effect of omission of cytarabine (ara-C) from treatment of newly diagnosed acute promyelocytic leukemia (APL), which allows administration of more anthracycline. PATIENTS AND METHODS Induction consisted of all-trans retinoic acid (ATRA) 45 mg/m2 daily until complete remission (CR) and idarubicin 12 mg/m2 daily for 4 days beginning on day 5 of ATRA. Patients in CR received two courses of idarubicin 12 mg/m2 daily for 3 days and then, until 2 years post-CR date, alternated three cycles of mercaptopurine, vincristine, methotrexate, and prednisone (POMP) with one cycle of idarubicin 12 mg/m2 daily for 2 days. Results in the 43 patients treated (41 with t(15;17) on standard or Southern analysis) were compared with those in 57 historic newly diagnosed APL patients given ara-C with either doxorubicin, amsacrine (AMSA), or daunorubicin without ATRA, using logistic and Cox regression to assess effects of non-treatment-related covariates on patient outcomes. RESULTS The CR rate in the current group was 77% (95% confidence interval [CI], 62% to 88%) and was not significantly different from the historic rate. In contrast, disease-free survival (DFS) in CR is superior in the current group (probability at 1 year 0.87; 95% CI, 0.73 to 1.0). This has translated into superior overall DFS for the current group (P = .03 adjusting for the predictive covariates initial WBC and platelet count; 1-year DFS probability 0.67; 95% CI, 0.52 to 0.82; median follow-up 102 weeks). The current treatment appears better both in patients with and without t(15; 17) on standard cytogenetic analysis. CONCLUSION Given the difficulties inherent in comparing sequential studies and recognizing the multiple differences in treatment between current and historic groups, our results suggest that a large randomized trial incorporating use of ATRA should assess the utility of omitting ara-C from treatment of newly diagnosed APL, thus allowing delivery of more anthracycline.