Effusion is a pathological accumulation of fluid in the abdominal, pleural, or pericardial cavity that results from one or more diseases, such as trauma, neoplasia, cardiovascular disease, metabolic disorders, and infectious and inflammatory diseases. Accumulation of fluid in body cavities results from an imbalance in fluid production and removal. Effusions are traditionally classified according to the concentration of total proteins and total nucleated cell count (TNCC). According to the above classification system, we distinguish between transudates (total proteins < 2,5 g/dl, TNCC < 1500 cells/μl), modified transudates (total proteins 2,5 – 7,5 g/dl, TNCC 1000 – 7000 cells/μl) and exudates (total proteins > 3,0 g/dl, TNCC > 7000 cells/μl). Laboratory examinations of effusions represent a diagnostic procedure that can aid in revealing the pathological process responsible for the accumulation of fluid in the body cavity (abdominal, pleural and pericardial). Collection and evaluation of fluid from body cavities may be a therapeutic as well as a diagnostic procedure. A modified classification of effusions on the basis of the possible cause includes the following groups: (1) protein-poor transudates, (2) protein-rich transudates, (3) septic exudates, (4) non-septic exudates, (5) effusions originating from ruptured blood vessels and visceral organs (hemorrhagic, lymphoragic or chylous effusions, uroperitoneum, and biliary peritonitis) and (6) effusions resulting from cell exfoliation. Very few proteomic studies of effusions have been conducted in veterinary medicine. Extracellular vesicles (EV), like exosomes are now recognized as important mediators of cell-cell communication. EV contain cell specific cargo, including specific sets of proteins, lipids, miRNA, mRNA and DNA and are also important regulators of cancer progression. Many studies indicate that exosomes such as flotillins play an important role in many biological processes like cell proliferation, apoptosis, adhesion and cell proliferation. Flotillins are overexpressed in various tumor processes and are closely related to tumor development, grade and metastasis. This study was performed in dogs with clinical signs of effusion in the pleural, pericardial and/or abdominal cavity. Dogs were admitted to the Clinic for Internal Medicine of the Faculty of Veterinary Medicine, University of Zagreb and referred for diagnostic and therapeutic procedures. The study included a total of 134 dogs, with clinical signs indicating the presence of effusions in body cavity or more cavities. Clinical examination, hematological and biochemical blood analysis, as well as electrocardiogram, x-ray and ultrasound findings served in the diagnosis of patients with effusions into body cavities. Dogs with clinical signs were divided into subgroups that consisted of non-tumor transudate, tumor transudate, non-tumor modified transudate, tumor modified transudate, non-septic non-tumor exudate, non-septic tumor exudate, septic non-tumor exudate and septic tumor exudate. Tumor related effusions were identified by finding of malignant cells in the effusion using cytological or histopathological examination. Patients with septic effusions were diagnosed based on the presence of intracellular bacteria in the smear and a positive bacterial culture. Patients with effusions which did not show any systemic disease, but were also negative for tumor, inflammatory or septic processes, were classified as idiopathic effusions. Abdominocentesis, thoracocentesis and pericardiocentesis were diagnostic as well therapeutic procedures for removal of large effusions. All body cavity centesis procedures were performed using aseptic technique, including proper hair removal and disinfection of the skin with chlorhexidine or povidone-iodine scrub before entering the body cavity. Blood for complete blood count was sampled in ethylenediaminetetraacetic acid (EDTA) tubes, and blood for biochemical tests in serum separating tubes containing gel. Additionally, the serum was separated by centrifugation at 1500 x g for 10 minutes, and was further used for biochemical tests. The effusion fluid was sampled in two tubes as well as blood samples (Vacutainer, Becton Dickinson, USA). Complete blood counts were performed on Animal Blood Counter (ABC, Horiba ABX, Diagnostics, Montpellier, France) using the manufacturer’s original solutions with the dog’s blood test device settings. Differential blood count was determined on May-Grünwald-Giemsa stained smears by counting 100 leukocytes under light microscope BX41 (Olympus, Japan). Biochemical parameters were determined in serum and in effusion on Architect c4000 biochemical analyzer (Abbott Laboratories, Illinois, USA) with original Abbott chemicals (Abbott Laboratories, Illinois, USA). Of the XI biochemical parameters, total proteins, albumin, glucose, aspartate aminotransferase (AST), lactate dehydrogenase (LDH) and canine C-reactive protein (canCRP) were determined. For effusion samples, hematocrit and total nucleated cell count were also determined. Total nucleated cell count was determined from the effusion collected in a tube with anticoagulant EDTA, using an automated hematology analyzer (Horiba ABX, Montpellier, France). Within 30 minutes after effusion sampling, smears were made, directly and/or from resuspended sediment of centrifuged effusion (depending on appearance, cell number and effusion density). The smear was stained by May-Grünwald-Giemsa method. In this study, a semiquantitative characterization of exosomal marker proteins flotillin 1, adiponectin, VCAM1 and CD63 was performed on selected serum and effusion samples of dogs with septic and tumor effusions. The exosomal proteins were detected by the Western blot method on the VMR Mini Vertical PAGE System (VWR International Ltd., Northern Ireland) and on the Electroblotting System (Mini Electroblotting System, Biovit, UK). Samples were loaded onto size exclusion chromatography columns (qEV, IZON, Oxford) to separate extracellular vesicles from soluble proteins. In addition to testing hypothesis of the existence of differences in the obtained values between the types of effusions, an analysis of other factors on the subject properties was performed. The statistical program R (version R 3.6.1., R DEVELOPMENT CORE TEAM, 2017) was used to check the significance and to include other factors that influenced the variability of properties in the model by the least squares method and the car package for analysis of variance. A corrected mean value (Least Square Means, LSM) was calculated for each property due to the unequal number of observations per impact classes using the emmeans package. The same package also tested the existence of statistically significant differences between the individual factors included in the model. Statistical significance of differences between arithmetic means in the studied dogs was tested by Bonferoni test. Of the study dogs 108 were males (81%) and 26 females (19%). As can be seen, a significant representation of males in relation to females is evident. Looking at the representation by body weight and size, dogs in the study group were divided into three breed groups according to the European Kennel Club and American Kennel Club classification, so that the population structure was: 20 dogs of small breeds (15%), 37 medium-sized (28%) ) and 77 large dog breeds (57%). The study included dogs with effusions in the body cavity, of which 81 (60,5%) were effusions in the abdominal cavity, 42 (31,3%) effusions in the pleural and 11 (8,2%) effusions in the pericardial cavity. The most common clinical signs in dogs with pleural effusions included dyspnea, respiratory distress, shrunken laying position with outstretched head and neck, open mouth breathing, rapid and pronounced abdominal breathing and in advanced cases, cyanosis was present. Dogs with abdominal effusion most often showed signs of lethargy, weakness, and abdominal tension, and in cases of peritonitis, abdominal pain was present. Clinical signs in dogs with pericardial effusion included weakness, lethargy, tachypnea, physical intolerance, collapse and cough. Modified transudates were the most common effusion in the body cavities of examined dogs (57,5 %). Neoplasia was a common cause of effusions in studied dogs. Adenocarcinoma and carcinoma of lungs and other visceral organs, lymphoma, hemangiosarcoma, histiocytic sarcoma and mesotelioma were the most common causes of tumor-related transudates. Non-tumor modified transudates were preceded by heart diseases, such as idiopathic dilated cardiomyopathy and endoventricular valve endocardiosis, while the remaining cases involved idiopathic effusions and trauma. The aim of this study was to discover practical diagnostic indicators in dogs with effusions in their body cavities, which would allow easier etiological classification of effusions in everyday clinical work. This could also help to predict the primary process outcome and determine the significance level of the ratio of selected biochemical parameters in serum and effusion (tumor and non-tumor etiologies, septic and non-septic etiologies). The application of an established algorithm for the classification of effusions based on the total nucleated cell count and total proteins in the effusion is a useful initial step in the diagnosis, which must be supplemented by additional biochemical and cytological parameters. The combined application of diagnostic procedures (biochemical analysis of effusion and serum, cytological analysis of effusion and advanced laboratory methods) was shown to be useful in determining the cause of effusion. The investigated laboratory indicators are of greater diagnostic importance in effusions than in the serum. TNCC showed the highest values in non-septic tumor exudates, both in dogs with pleural and abdominal effusions. The lowest TNCC was found in transudates. Total proteins showed similar values in sera of dogs from all investigated groups of effusions. Differences found in the values of total proteins in effusion samples were significant only for abdominal effusions. Glucose concentration in dogs with abdominal effusions is a a useful indicator for differentiation of septic tumor from septic non-tumor exudates, as well as non-tumor modified transudates from tumor modified transudates. Determination of lactate dehydrogenase (LDH) concentration in dogs with pleural and XIII abdominal effusions facilitates the etiological classification of effusions. LDH ratio (effusion/serum) showed to be the best laboratory indicator for distinguishing tumor from non-tumor and septic from non-septic causes both in pleural and abdominal effusions. Glucose ratio (effusion/serum) and CRP ratio (effusion/serum) in abdominal effusion have shown to be good indicators for distinguishing the causes of individual effusions (tumor and non-tumor etiology; septic and non-septic etiology). Cytology was shown to be a valuable diagnostic procedure in the investigated population of dogs with effusions, especially for detection of malignant cells, inflammatory cells and microorganisms (mostly intracellular bacteria). In this work, sensitivity of cytology for identification of septic peritonitis was 95%. Most of the pericardial effusions were hemorrhagic and were linked to the presence of hemangiosarcoma, idiopathic pericardial effusion and mesotelioma. The most common cell types found in pericardial effusions were macrophages, neutrophils and reactive mesotelial cells. Cytological examination was hampered by the proliferation of mesotelial cells which can mimic malignant cells. In addition to the research of laboratory indicators (most often including cytological and biochemical tests) in effusions and sera of the entire studied population of dogs, part of the work is focused on expanding knowledge about other biomarkers, such as exosomes. Their potential as an auxiliary tool in the diagnosis of the most common and diagnostically most challenging tumor and septic diseases in dogs with effusions into their body cavities was investigated. In previous studies, little attention has been paid to the study of exosomes in the body cavities of dogs, so in order to expand this knowledge, exosomal proteins such as flotillin 1, vascular cell adhesion molecule 1, adiponectin and CD63 protein were investigated in serum and effusions to discover the molecules by which it is possible to distinguish the causes of processes, especially in distinguishing tumor from septic processes. Expression of selected exosomal proteins such as flotillin 1, adiponectin and CD63 was found useful for distinguishing tumor-related effusions from septic effusions. Flotillin 1 and CD63 are overexpressed in both tumor and septic processes. The expression of flotillin 1 and adiponectin are significantly higher in the serum of dogs with tumors and may serve to distinguish tumor from septic processes in dogs. VCAM1 as a molecule which expression is closely related to tumor angiogenesis, metastases, severe inflammation and chronic conditions of some diseases, was not found as a marker of increased expression in septic and tumor conditions that cause effusions in the body cavities of dogs. Exosomes play a significant role as mediators in sepsis and in the spread of cancer cells. Early diagnosis and access to treatment can prevent serious complications in dogs with existing effusions in body cavities, regardless of various etiological factors. Therefore, it is crucial to identify biomarkers that can be applied in clinical practice. Using selected laboratory parameters and their ratios and increased expression of exosomal proteins in serum and effusions into the body cavities of dogs, effusions caused by the tumor process can be distinguished from effusions caused by septic causes.