The blast population in acute myeloblastic leukemia (AML) contains cells capable of forming blast-cell colonies in culture. The purpose of this study was to measure the effects of hydrocortisone on this process, using two end-points. First, we measured the effects of increasing concentrations of hydrocortisone on the primary plating efficiency of T-lymphocyte-depleted blast cell preparations from AML peripheral blood. Second, colonies forming in the presence or absence of the hormone were pooled and replated; changes in the plating efficiencies (secondary plating efficiency or PE2) of these suspensions reflected the effect of the hormone on blast progenitor self-renewal. For comparison, we measured the hydrocortisone dose response curves for normal granulopoietic and T-lymphocyte colony-formation. The latter showed little individual variation; T-lymphocyte colony-formation was regularly sensitive to the hormone while granulopoietic colony-formation was resistant. In contrast, wide variations were found in the hydrocortisone dose response curve for blast from 24 patients with AML (FAB 1-6). A significant association was found between successful remission induction and resistance to hydrocortisone in 24 treated patients. The association was maintained when the data was stratified by other risk factors, including PE2 and the presence of blasts bearing immunologically-defined markers of more than one differentiation lineage (lineage infidelity). We propose that sensitivity to hydrocortisone may reflect the passage of blast cells through lymphopoiesis-associated components of differentiation programs. From this point of view, the poor prognosis associated with sensitivity of blast progenitors to hydrocortisone may be similar to the response-failure of patients whose blasts exhibit lineaged infidelity when tested with immunological procedures.