Photodegradation of an industrial Azo dye C.I Basic Red 46, was examined in a semi-pilot scale prototype solar photoreactor under solar radiation. In our study, photodegradation of the dye was optimized using Response Surface Methodology (RSM) based on Box-Wilson approach. The Artificial Neural Network (ANN) was used to establish suitable modeling and optimal conditions for the Solar UV/Immobilized-TiO2 process in order to evaluate the individual effects of three factors that independently affect the effectiveness of the photodegradation process: (1) initial concentration of the dye, (2) pH, and (3) flow rate. The RSM was in good agreement with the prediction model (R2Dec = 0.95); meanwhile, the ANN approach revealed that the predicated model fit perfectly with the experimental data to yield the highest value of R2 = 0.999. The effects of these three factors could be estimated from a second-order polynomial equation, and the optimal parameters of photodegradation consisted of three main parameters: (1) initial concentration of colorant 10.65 mg.L−1, (2) pH 10.82, and (3) rate of fluid flow of 852 L h−1. The decolorization removal efficiency under these optimal conditions was 99%.