We retrospectively reviewed our experience with the fine-needle aspiration biopsy (FNAB) diagnosis of primary and recurrent lymphoma to assess the ability of cytomorphology with and without ancillary flow cytometry (FCM) analysis to diagnose and subclassify these tumors according to the Revised European-American Lymphoma/World Health Organization classifications. We reviewed 139 consecutive FNABS of 84 primary and 55 recurrent lymphomas. FCM was successful in 105 (75%) cases. The overall results, including cases without FCM, included 93/139 (67%) true positive, 7 (5%) false negative, and 39 indeterminate (27 [19%] suspicious and 12 [9%] atypical) diagnoses of lymphoma. In cases with FCM, there were 80/105 (77%) true positive, no false negative, and 25 indeterminate diagnoses (15 [14%] suspicious and 10 [9%] atypical). The overall results of the 84 primary lymphomas were 55 (67%) true positive, 5 (5%) false negative, and 24 indeterminate (14[16%] suspicious and 10 [12%] atypical) diagnoses for lymphoma. Of the 68 primary lymphomas analyzed with FCM, 50 [74%] were true positives, and 28 were indeterminate (11 [16%] suspicious and 7 [10%] atypical). There were no false negatives. Diagnostic accuracy varied among lymphoma subtypes. Subclassification of the positive cases were initially conclusive in only 55/93 cases (59%). However, a retrospective review of the morphologic together with FCM data in 15 of the 23 unclassified cases improved the overall subclassification of positive cases to 77%. Subclassification was best in small lymphocytic lymphoma/chronic lymphocytic leukemia, lymphoplasmacytic lymphoma, Burkitt's lymphoma, mantle cell lymphoma, and plasmacytoma (all 100%). Subclassification was poor in marginal-zone lymphoma (33%), and initially as well in diffuse large B-cell lymphoma (62%), but it improved on review (95%), as did subclassification of follicular lymphoma (77 to 100% on review). Hodgkin's disease was recognized as malignant in only 44% of the cases (7/16) and was classified as such based on morphology alone. This review of our early efforts to diagnose and subclassify lymphoma with FNAB and FCM indicates that although a diagnosis and proper subclassification of lymphoma can be made with certainty in the majority of cases, recurrent or primary, it requires close coordination of cytomorphology and immunophenotyping data, which often comes with close cooperation of cytopathologists and hematopathologists. A mere cytological diagnosis of positive for lymphoma is no longer acceptable if FNAB is to become an independent diagnostic tool for lymphoma.