In the cooperative study AML-IGCI-84 27 children with AML (FAB Ml 7 × M2 4 × M3 1 x M4 6 × and M5 8 ×; 1 megakaryocytic leukemia) have been treated. The median initial white blood cell count was 18.0 G/l (range 1.8-1,350.0 G/l). 1 or 2 courses of induction therapy were used: II (aclacinomycin-A (ACLA-A), VP-16 and ARA-C) and 12 (daunorubicin (DNR), VP-16, and ARA-C). 12 was used only if bone marrow contained > 5 % blast cells on day 21.12 and consolidation treatment were identical with the current AML-BFM-83 protocol. 3 deaths before day 21 occurred (2 cerebral hemorrhages, 1. septicemia). 24 patients were evaluable for response, 20 (83.3%) achieved CR, 16 (66.7%) by II, 4 after 12. 4 patients never reached CR, 3 of them had a PR after II. M5 patients did badly (2 early deaths, 2. PR, 4 CR). All patients without CR after II received the whole AML-BFM-83 protocol. Comparison of the results of the 2 studies revealed a similar CR rate for II (our patients) and 12 (BFM data): 80.0% vs. 82.2% (calculated for patients who ever reached CR). CR was reached before consolidation in all our CR patients compared to 82.2% of BFM patients. Early CR may be of long term prognostic significance. Cardiotoxicity of induction may be reduced by substitution of DNR by ACLA-A.