We studied the impact of cytogenetics and kind of induction/consolidation therapy on 848 adult acute myeloid leukemia (AML) patients (age 15-83). The patients received three types of induction/consolidation regimen: standard (daunorubicin and cytosine arabinoside (3/7); two cycles); intensive (idarubicin, cytosine arabinoside and etoposide (ICE), plus mitoxantrone and intermediate-dose Ara-C (NOVIA)); and low-dose (low-dose cytosine arabinoside). CR patients under 60 years of age, if an HLA-identical donor was available received allogeneic stem cell transplantation (allo-SCT); otherwise, as part of the program, they underwent autologous (auto)-SCT. CR rates significantly associated with 'favorable' (inv(16), t(8;21)), 'intermediate' ('no abnormality', abn(11q23), +8, del(7q)) and 'unfavorable' (del (5q), -7, abn(3)(q21q26), t(6;9), 'complex' (more than three unrelated cytogenetic abnormalities)) karyotypes (88% vs 65% vs 36%, respectively; P = 0.0001). These trends were confirmed in all age groups. On therapeutic grounds, intensive induction did not determine significant increases of CR rates in any of the considered groups, with respect to standard induction. Low-dose induction was associated with significantly lower CR rates. Considering disease-free survival (DFS), multivariate analysis of the factors examined (including karyotype grouping) showed that only age > 60 years significantly affected outcome. However, in cases where intensive induction was adopted, 'favorable' karyotype was significantly related to longer DFS (P = 0.04). This was mainly due to the favorable outcome of t(8;21) patients treated with intensive induction. Patients receiving allo-SCT had significantly longer DFS (P = 0.005); in particular, allo-SCT significantly improved DFS in the 'favorable' and 'intermediate' groups (P = 0.04 and P = 0.048, respectively). In conclusion our study could provide some guidelines for AML therapy: (1) patients in the 'favorable' karyotype group seem to have a longer DFS when treated with an intensive induction/consolidation regimen, adopted before auto-SCT instead of standard induction; this underlines the importance of reinforcement of chemotherapy, not necessarily based on repeated high-dose AraC cycles. Allo-SCT, independently of induction/consolidation therapy, should be considered an alternative treatment; (2) patients in the 'intermediate' karyotype group should receive allo-SCT; (3) patients in the 'unfavorable' karyotype group should be treated using investigational chemotherapy, considering that even allo-SCT cannot provide a significantly longer DFS, but only a trend to a better prognosis.