Background: Two classifications are used to categorize cavitary effusions using total nucleated cell count (TNCC): protein concentration and pathophysiology of its formation. The aims of the present study were to evaluate the correlation between the TNCC values of cavitary effusions obtained in the automatic and the manual method, and also evaluating the classification methodology. Materials, Methods & Results: Cavitary effusions were analyzed for physical, chemical and cytological aspects, as well as manual and automatic cell counts for the correlation between the traditional methods and those suggested by Stockham & Scott. Bland-Altman regression and Spearman correlation analysis were performed. Of the total, 44 were abdominal effusions (73.3%), 15 thoracic (25%) and 1 pericardial (1.7%). According to the traditional classification, most of the effusions were classified as modified transudates (40%) and according to the classification of Stockham and Scott, as transudates poor in protein (31.7%). The correlation between cell counting techniques between pure, modified and exudate transudates was 0.94, 0.97 and 0.94, respectively, indicating an excellent correlation between the parameters (p = 0.95%).Discussion: Considering the concentration of proteins and CCNT, the effusions classified as modified transudate were mainly caused by neoplastic processes (carcinomas/adenocarcinomas), since there are several mechanisms of their formation, such as large variation of protein concentration. According to the Stockham & Scott classification a unique classification is considered for exfoliative neoplastic effusions, the variation of the protein concentration of the effusion does not alter its classification. In neoplastic effusions, classified as exudates, lymphomas were the most prevalent, and hypercellularity (approximately 150,000 cells / μL) allowed this classification. When considering low-protein transudates, the findings related to low concentrations did not differ much from the traditional classification. In the ruptures of viscera and vessels, the hemorrhagic ones were the most frequent, thus, the cytological diagnosis is essential, since it can give information about the contamination with blood during the collection. Most of these were due to neoplasia as the underlying cause. A case of chylotorax was diagnosed by comparing cholesterol and triglyceride values of effusion and serum. In cases of uroperitoneum, the presence of urine in the abdominal cavity promotes the dilution of the fluid from the cavity, being initially classified as pure transudate and, with its permanence in the cavity, increasing the CCNT, becomes an exudate. As in cases of exfoliative neoplastic effusions, the classification of the uroperitoneum, according to Stockham & Scott, is classified directly into effusion due to rupture of the viscera, giving a quick and clear diagnosis. According to Stockham & Scott, cases classified as nonseptic exudates (n = 3), two of which resulted from feline infectious peritonitis (PIF). The effusive form of PIF presents with accumulations of fluid in the abdomen, having an inflammatory character, but according to the traditional classification, they enter the category of modified transudates, because, despite containing protein concentrations close to or above the serum level, they present a CCNT lower than an exudate. Cavitary effusions were classified as septic exudates when intracellular bacteria were present and in the present study, two effusions were classified as such in two patients, one with septic peritonitis and in the other the final diagnosis was not found. The high values of Spearman correlation coefficients found when comparing the automatic counts with the manual demonstrate that there is an excellent correlation between the methods and, the Bland-Altman test showed significant agreement between them.