Interphase fluorescence in situ hybridization (I-FISH) for the BCR-ABL translocation performed on peripheral blood (PB) white cells has been suggested as a surrogate for conventional bone marrow (BM) cytogenetics for monitoring patients with chronic myeloid leukemia (CML). I-FISH is faster, less costly, and does not require BM aspiration. For patients treated with interferon-alpha (IFN), a good correlation between the two methods has been demonstrated in several though not all studies. However, imatinib mesylate (STI571) has largely replaced IFN as the standard drug treatment for CML, raising the question if the results obtained in IFN-treated patients are applicable to patients on imatinib. We therefore compared the two methods in patients on imatinib and patients on other therapies, mainly IFN (collectively referred to as nonimatinib therapies). Our results demonstrate that the correlation between I-FISH and cytogenetics is much weaker in patients on imatinib than in patients on nonimatinib therapies. Correction of the I-FISH values for the proportion of lymphocytes barely improved the correlation, probably as a result of unpredictable proportions of Philadelphia-positive B cells. By contrast, I-FISH of PB neutrophils was much better correlated with BM cytogenetics. We conclude that I-FISH on unselected PB white cells is not suitable for monitoring patients on imatinib.