The identification of non-hematopoietic cells in effusions is a diagnostic challenge in cytology. Biopsies from mesothelium or primary lesions are infrequently performed in clinical settings and immunochemistry on smears or immunohistochemistry on cell blocks are the most common ancillary test to refine the cytological diagnosis. Cavitary effusions are an ideal matrix for flow cytometry and the availability of a cytometric panel to describe non-hematopoietic cells would represent a useful tool. Here we present the results of the flow cytometric and immunohistochemical determination of cytokeratin (CK), vimentin (VIM) and desmin (DES) in 36 canine effusions. The concordance between the two methods was perfect for CK (100%), substantial for VIM (77.8%), and almost perfect for DES (97.2%). The panel was interpreted to define the epithelial (CK+VIM-DES-), mesothelial (CK+VIM+DES+), or mesenchymal (CK-VIM+DES-) origin of the cells. Unexpected profiles were considered doubtful and observed patterns were individually discussed. The concordance of the panel interpretation between two methods was 75%. The evaluation of discordant and doubtful cases suggests a lower sensitivity of flow cytometry in detecting VIM expression and revealed a high frequency of VIM+ epithelial cells, variable expression of VIM in mesothelial cells, and an important role of DES in excluding an epithelial origin when positive. Multicentric studies based on histopathological diagnoses are necessary to confirm these findings and evaluate the diagnostic utility of the panel to refine cytological diagnosis. Our results show that flow cytometry can be a timesaving alternative to IHC on cell blocks in clinical settings to detect CK, VIM and DES expression. The interpretation of the panel is similar in most cases; however, occasional discordant results, particularly for VIM, may occur.