ATRA combined to anthracycline based chemotherapy is the reference treatment of newly diagnosed APL and a role for maintenance treatment is supported by many studies, but some controversies remain, like the role of AraC. Methods: We performed a joint analysis of pts 2 ), and maintenance with intermittent ATRA and continuous 6MP + MTX during 2 years. Pts with high (WBC > 10000/mm3) and intermediate (WBC 2 ), vs low risk pts (no consolidation ATRA, cumulative IDA: 80 mg/m 2 ). No AraC was administered in LPA99 trial. For APL 2000, we restricted the analysis to treatment groups that received AraC: pts with low or intermediate Sanz score received induction with ATRA and DNR (60 mg/m 2 /d x 3) + AraC (200 mg/m 2 /d x 7) followed by consolidation with a similar course and a final DNR (45 mg/m 2 /d x 3) + AraC (1 g/m 2 /12h x 8) course and the same maintenance as in LPA99. High risk patients received the same treatment, but with intrathecal CNS prophylaxis and, in pts 2 /12h x 8 during the last course. Median follow up was 30 and 22 months in LPA99 (412 pts) and APL 2000 (180 pts), respectively. Comparisons were adjusted on age, gender, WBC and platelet counts. Results: In low and intermediate risk patients (308 LPA99 pts, 97 APL 2000 pts), the CR rates, 2 year cumulative incidence of relapse (CIR), EFS and survival were 96% and 99% (p = 0.2), 2.5% and 6.9% (p = 0.05), 93% and 91% (p = 0.6), 95% and 97% (p = 0.3) in LPA99 and APL 2000 trials, respectively. In high risk patients (104 LPA99 pts, 83 APL 2000 pts), the CR rate, 2 year CIR, EFS and survival were 84% and 95% (p = 0.02), 16.0% and 3.3% (p = 0.06), 69.2% and 88.0% (p = 0.01), 82.4% and 91.2% (p = 0.04) in LPA 99 and APL 2000, respectively. Four relapses in high risk pts involved the CNS in LPA99, vs. none in APL 2000. Three and 5 pts in LPA99 and APL 2000 respectively, died in CR, from treatment toxicity. Conclusion: This analysis suggests than in pts with WBC