Development of drug resistance is a major problem in attempts to control the growth of neoplastic cell populations. The resistance can he either inherent or acquired by an exposure to a chemotherapeutic drug. The available models for study of these phenomena have not led to major improvements in the therapy for human cancers. Therefore, in order to develop a new model for such studies, we have exposed human myeloid leukemia cells HL60 to increasing concentrations of 1,25-dihydroxyvitamin D3 (1,25D3) and characterized the emerging new phenotypes of these cells over a period of 4 years. During the stepwise development of resistance only cells which did not adhere to the flask walls were passaged. Beginning at 30 nM 1,25D3 the sublines became resistant to the differentiation-inducing and growth-retarding properties of 1,25D3 even at 400 nM. Also, their growth rates in 1,25D3-free media increased. In addition, beginning at 40 nM 1,25D3 the sublines acquired resistance to 5-beta-D-arabinocytosine (araC) due to the lack of expression of the deoxycytidine kinase gene. The araC-resistant sublines were also near-tetraploid, as judged by their DNA content. When grown in 1,25D3-free long-term culture the phenotype was essentially stable. The development of cross-resistance to araC during growth in the presence of an unrelated compound (i.e., 1,25D3) shows that in some instances an apparently inherent drug resistance may in fact he due to a metabolic defect resulting from an exposure to another agent.