The oxygen transfer process through the air-water interface in free surface flows can be numerically quantified by the coefficient of surface reoxygenation . There are many empirical predictive equations in the scientific literature that consider only two water body hydraulic parameters, such as flow velocity and depth. Some studies have demonstrated that 14 parameters are important in the oxygen transfer process through the air-water interface. Since only two parameters were considered, the predictive equations available in the literature commonly predict very different results for the same hydraulic conditions. The present work aims at the generation of new data in order to contribute to the understanding of the reoxygenation process, including one more parameter in the analysis: channel bed roughness. The study of the influence of the water velocity, depth and roughness in hydraulic channel flows on was performed aiming at developing a predictive semi-empirical model. A hydraulic recirculating channel was used to simulate the reoxygenation process in water bodies. The results show that, for a certain range of values, the roughness of the bed significantly interferes in the reoxygenation process. The model developed can be used to predict the in flows with the same hydrodynamic and geometric characteristics established in this study.