Ninety-three (93) cases of acute leukemia were assessed using flow cytometry and cytochemistry and assigned to one of four categories: myeloid, lymphoid, biphenotypic, and non-diagnostic. In leukemias designated as ALL or AML by both methodologies, there was lineage agreement in all but 3 of 71 cases (95.8%). However, when nondiagnostic or biphenotypic diagnoses made by either methodology were included, complete agreement occurred in only 77.4% of cases. Of 37 cases designated myeloid origin by flow cytometry, 33 (89.2%) were read as myeloid by cytochemistry. The four discordant diagnosis were read as lymphoid (2) or as non-diagnostic (2). Eighty percent of lymphoid leukemias were diagnosed as such by both flow cytometry and cytochemistry; one early B cell ALL was diagnosed as myeloid and 8 as non-diagnostic. Fifty percent (50%) of flow cytometry defined T-cell ALL were considered non-diagnostic by cytochemistry as compared to 17% of the total ALL group. Of the remaining four designated non-T cell ALL by flow cytometry and non-diagnostic by cytochemistry, three were read by flow cytometry to be standard pre-B ALL and one an early B-cell ALL. Only 2/9 leukemias considered biphenotypic by flow were identified as such by cytochemistry. Given (1) the potential importance of non-lineage expression in the prognosis of myeloid and lymphoid leukemias, (2) cytochemistry's impaired ability to diagnose biphenotypic, T-cell, and promyelocytic leukemias, and (3) the increased costs incurred in diagnosis when both modalities are used, perhaps it is time to re-examine the utility of performing both flow cytometry and cytochemistry as initial testing for leukemia categorization.