<h3>Background</h3> Rheumatoid arthritis (RA) is an inflammatory disease with autoimmune manifestations characterized by chronic inflammation of the joints associated with systemic complications [1]. Systemic inflammation of RA causes metabolic changes and deregulation of leptin, adiponectin and insulin, leading to anorexia, changes in normal food intake and weight loss [2]. However, observational with human subjects have many potential bias and animal models can be a good alternative for studies on the effect of chronic arthritis in metabolic changes and feeding preferences. <h3>Objectives</h3> To evaluate the feeding preferences of rats with type II collagen induced arthritis (CIA). <h3>Methods</h3> Female Wistar rats were separated in two groups: control (CO) and collagen-induced arthritis (CIA) performed accordingly to Rosloniec et al. [3] with modifications (without booster). Arthritis clinical score (0-16) of animals was registered three times a week for analysis of the development of the disease [3]. All animals were offered four different diets at the same time: control diet (standard ration - CD), high calorie diet (addition of 5% of saccharose – HCD), high protein diet (addition of 5% of albumin – HPD) and high fat diet (addition of 5% of soybean oil – HFD). The animals and ration leftovers were weighted every three days for 14 days before and 25 days after the induction of arthritis. Comparison between groups was performed by two-way ANOVA followed by Tukey post-hoc test. <h3>Results</h3> The first signs of arthritis appeared by 14 days after induction (score 2.7±3.1) and the inflammatory peak appeared to be around day 17 to day 22 (score 9.8±4.1). Food consumption of control and CIA rats before the onset of arthritis was: 45% of HCD (51.9±6.4 g), 36% of HFD (41.6±9.7 g), 13% of HPD (15.3±5.1 g) and 7% of CD (7.5±5.8 g). During the peak of arthritis, CIA rats had a significant weight loss (-14.7±8.5 g) while CO group maintained their weight (-0.2±4.6 g, p<.01). At the same time, CIA had reduced total food intake (72.3 ± 8.6 g) compared to CO (107.3±5.7 g, p<.01). Besides that, CIA changed the food preference, increasing the intake of HPD to 20% (16.3±5.3 g, p<.05) and reducing the ingestion of the HCD to 27% (22.6±7.8 g, p<.05) compared to CO (10.0±1.5 g and 58.3±21.3 g, respectively). <h3>Conclusions</h3> CIA animals demonstrated changes in their food intake, especially during the inflammatory peak of arthritis, with concomitant reduction in weight. These changes include decreased total food consumption and different food preferences, like increased intake of high-protein diet and reduced intake of high-calorie diet, results never reported before. There is still needed more studies about the feeding preferences and intake, but CIA model could be a good model to study altered food consumption caused by arthritis and hereafter correlate these findings with alterations in RA patients. <h3>References</h3> Mc Innes IB, et al. Engl J Med, 2011. 365(23):2205-19. Argilés JM, et al. J. Biocel, 2005. 37(10):2036-46. Rosloniec EF, et al. Curr Protoc Immunol, 2010. Chapter 15: Unit 15.5 1-25. <h3>Acknowledgements</h3> FIPE-HCPA, CAPES, CNPq <h3>Disclosure of Interest</h3> None Declared