When a certain load of pollutants is dumped in urban surface waters, the natural hydrological parameters are altered and a consequent dispersion of a contaminant occurs. Water self-purification is a natural recovery mechanism that aims to restore the natural parameters of water quality. In the municipality of Santarém, the Urumari micro-watershed, one of the main urban streams and a tributary of the Amazon river, suffers from increasing degradation. Thus, it is necessary to systematically monitor the water quality of the micro-watershed, through mathematical modeling, in order to identify environmental impairment and propose mitigating tools that facilitate self-purification capacity. The aim of this work was to evaluate the degree of disturbance of the Urumari micro-watershed using Streeter-Phelps mathematical models and the Enhanced DO-BOD model. The software Curve Expert 1.4 was used to adjust the coefficients of the mathematical models, using the Levenberg-Marquardt nonlinear function adjustment method. The complete mathematical model of DO and BOD considered the parcels of punctual deficit, microbial BOD, nitrification, sediment demand, photosynthesis and respiration. Three coefficients of the Streeter-Phelps model and six coefficients of the Enhanced DO-BOD model were adjusted according to the experimental data of dissolved oxygen and water travel time. The mathematical models satisfactorily correlated the experimental data. The values of the correlation coefficient, mean absolute deviations and quadratic deviations are for the Streeter-Phelps model and the Enhanced DO-BOD model, respectively. The Streeter-Phelps model presented the best correlation to the experimental data with the values of R2 equal to 0.83; mean absolute deviation equal to 0.56 and quadratic deviations equal to 4.98.