ETHNOPHARMACOLOGICAL RELEVANCE: Dilodendron bipinnatum (Sapindaceae) stem bark decoction and macerate were used to treat uterine inflammation, pain in general, dermatitis and bone fractures. These homemade preparations also have diuretic, stimulant, expectorants and sedative effects and are effective in treating worm infections in the Brazilian Pantanal population. Our previous research confirmed the anti-inflammatory activity of the hydroethanolic extract of inner stem bark of D. bipinnatum (HEDb). AIM: This work aimed to investigate the efficacy of HEDb in ameliorating experimental colitis in rats and to elucidate the possible mechanisms involved in the anti-ulcerative colitis properties of HEDb in rats and Caco-2 cell line. MATERIALS AND METHODS: The effects on cell viability, IL-8 and TNF-α in human colon adenocarcinoma (Caco-2) were determined by flow cytometer and ELISA. Wistar rats (n = 6-7) were orally gavaged with, vehicle (0.9% saline), HEDb at doses of 20, 100 or 500 mg/kg, or mesalazine at a dose of 500 mg/kg, at 48, 24 and 1 h prior to the administration of trinitrobenzene sulfonic acid via rectal administration to induce colitis. The anti-inflammatory effects of HEDb were assessed macroscopically, by myeloperoxidase (MPO) activity and for glutathione (GSH) concentration in the colon. Additionally, colonic histopathological analyses of UC severity were conducted by different staining methods (H&E, PAS and toluidine blue). Pro-inflammatory cytokines TNF-α and IL-1β were quantified in colonic tissue by ELISA and colonic expressions of COX-2 and IL-17 were analyzed by western blotting. RESULTS: HEDb was shown to be non-cytotoxic with mean viability of 80% in Caco-2 cells. HEDb pre-treatments of 1, 5 or 20 μg/mL significantly reduced TNF-α production in Caco-2 cells by 21.8% (p < 0.05), 60.5 and 82.1% (p < 0.001) respectively following LPS treatment compared to LPS alone. However, no change in IL-8 production was observed. HEDb pre-treatment of rats subjected to TNBS significantly (p < 0.001) reduced colonic lesion score. Higher doses (100 and 500 mg/kg) caused a sharp downregulation of haemorrhagic damage, leukocyte infiltration, edema and restoration of mucus production. Moreover, mast cell degranulation was inhibited. Colonic MPO activity was reduced following all doses of HEDb, reaching 51.1% ± 1.51 (p < 0.05) with the highest dose. GSH concentration was restored by 58% and 70% following 100 and 500 mg/kg of HEDb, respectively. The oral treatment of HEDb at doses 20, 100 and 500 mg/kg decreased the concentrations of TNF-α and IL-1β at all doses in comparison to vehicle treated control. In addition, HEDb inhibited the COX-2 and IL-17 expressions with maximal effect at 500 mg/kg (60.3% and 65% respectively; p < 0.001). In all trials, the effect of HEDb at all doses being 20, 100 and 500 mg/kg was statistically comparable to mesalazine (500 mg/kg). CONCLUSIONS: HEDb reduces colonic damage in the TNBS colitis model and relieves oxidative and inflammatory events, at least in part, by increasing mucus production, reducing leukocyte migration and reducing TNF-α (in vivo and in vitro), IL-1β, IL-17 and COX-2 expression. Therefore, HEDb requires further investigation as a candidate for treating IBD.