Recent studies indicate that a rare population of primitive quiescent BCR-ABL(+) cells are innately insensitive to imatinib mesylate (IM) and persist after IM therapy of patients with chronic myeloid leukemia (CML). New approaches to the eradication of these cells are therefore likely to be crucial to the development of curative therapies for CML. We have now found that Ara-C, LY294002 (a PI-3 (phosphatidylinositol-3' kinase) kinase inhibitor), 17AAG (a heat-shock protein (HSP)-90 antagonist) and lonafarnib (a farnesyltransfease inhibitor) all enhance the toxicity of IM on K562 cells and on the total CD34(+) leukemic cell population from chronic phase CML patients. However, for quiescent CD34(+) leukemic cells, this was achieved only by concomitant exposure of the cells to lonafarnib. Ara-C or LY294002 alone blocked the proliferation of these cells but did not kill them, and Ara-C, LY294002 or 17AAG in combination with IM enhanced the cytostatic effect of IM but did not prevent the subsequent regrowth of the surviving leukemic cells. These studies demonstrate the importance of in vitro testing of novel agents on the subset of primary leukemic cells most likely to determine long-term treatment outcomes in vivo.