To illustrate methodological issues, we compared donor vs no-donor to transplant vs no-transplant comparisons in a cohort of 107 patients aged 50 years with adverse karyotype AML in first CR. Adverse karyotypes were defined as -7, del(7q), -5, del(5q), t(9;22), 11q23, 3q26 or complex abnormalities. Mantel-Byar estimations and hematopoietic SCT (HSCT) as a time-dependent variable were used to compare transplant vs no-transplant cumulative incidence of relapse (CIR), relapse-free survival (RFS) and OS. In all, 52 patients had a sibling donor, but only 35 of them were transplanted in first CR, whereas 9 patients received HSCT from alternative stem cell sources. Donor-based analysis showed lower CIR in the donor group, not translating in prolonged RFS or OS. Conversely, transplant-based analysis showed that HSCT in the first CR improved the three CIR (multivariate hazard ratio (HR), 0.31; P<0.001), RFS (multivariate HR, 0.57; P=0.047) and OS (multivariate HR, 0.54; P=0.03) endpoints. At 5 years, OS was estimated at 33% in transplanted vs 18% in non-transplanted patients. The positive effect of HSCT was more pronounced in patients aged 35 years and/or in those transplanted in the more recent years. These results confirm that HSCT is likely the best curative option in younger patients with adverse karyotype AML.