Attempts to reproduce inflammatory colitis have created many experimental models. Since the pioneer work of Morris et al. (1989), trinitrobenzenosulfonic (TNBS) solutions have been used with different dosages. The aims of this work were standardize the induction of colitis, evaluate the clinical and intestinal effects of different doses and verify the reproducibility of the intestinal inflammatory process. Wistar rats were inoculated endo rectally with 2.5 ml solutions of different concentrations of TNBS and ethanol, and the rats were sacrificed after 14 days. According to the solution concentrations of TNBS (mg) and ethanol (%), six groups of animals were established: Control (saline), 30 mg/30%, 30 mg/40%, 30 mg/50%, 50 mg/10% and 50 mg/30%. Statistical analysis of food ingestion showed no differences between groups (p = 0.247). The 30/50 group presented greater weight loss when compared to 50/10 and 50/30 groups (p = 0.012). Groups 30/50 and 30/40 showed greater degrees of macroscopic lesion than control and 50/10 group (p < 0.05). Histologic lesion was not uniform to all rats regardless of the solution employed. Group 50/10 presented the less severe histologic alterations; on the other hand, 30/40 and 30/50 groups had important changes on mucosal thickness, on vascularization and ulceration. The authors conclude that experimental colitis with TNBS 1) cause intestinal lesions that are not uniform to all animals, although they may be reproduced in many of them; 2) with the same doses of TNBS, the increase in ethanol concentrations leads to a greater inflammatory process, intestinal thickness, vascularization, abscess formation and intestinal ulceration; 3) 50/10 and 50/30 solutions make less severe lesions when compared to 30/40 and 30/50 solutions; 4) 30/50 solution was the best one, as it produces inflammation 90% of the animals, increases in wall thickness in 50%, abscess in 70% and ulceration in 38%.