The effectiveness of the solar-driven photo-Fenton process in treating wastewater contaminated with aminosilicone compounds has been evaluated in a parabolic trough reactor under variable weather conditions. On sunny days, after 3 h of irradiation, more than 80% of the initial COD has been removed. The degradation generates compounds that might be readily biodegradable and/or a solid phase that is easily separated by mechanical means. An important interaction between manipulated [H2O2 and Fe(II) concentrations] and nonmanipulated (direct and diffuse components of solar radiation) variables was detected. Therefore, degradation was possible even on cloudy days, provided that the H2O2 and Fe(II) concentrations are conveniently selected. The neural network technique is an effective, simple approach to successfully modeling the solar-driven photo-Fenton degradation. The model might therefore be useful in process optimization, as well as in the design and scale-up of solar reactors for industrial application.