Fluorescence in situ hybridization for abnormalities common to myelodysplastic syndrome (MDS FISH) is often used with traditional karyotype in the diagnosis and monitoring of myeloid neoplasms. However, its value in these roles has been questioned. To evaluate its utility, we compared MDS FISH results with karyotype in 544 bone marrow specimens obtained for diagnosis (180 cases) or follow-up (364 cases) of myeloid neoplasia. We found excellent concordance between FISH and karyotype, such that FISH is rarely abnormal (1.7% at diagnosis and 3.0% at follow-up) in cases with normal karyotype. Even in the rare discordant cases, the abnormal FISH has little or no clinical value. Thus, we propose that this test should be limited to cases with inadequate karyotype only. Such guidelines could result in significant cost savings with no impact on patient diagnosis.