A kinetic study was accomplished to describe the photocatalytic production of hydrogen in liquid phase. A reaction mechanism and a kinetic model were proposed to predict the rate of hydrogen production, which is a function of light intensity, catalyst loading, substrate concentration, and time. To assess the capability of the proposed model, glycerol and ethanol were selected as representative hydrogen sources (substrates). The experimental data performed under different operating conditions, based on Box-Behnken experimental design, were used to train the developed kinetic model, optimize the parameters using genetic algorithms and check its accuracy. The analysis confirms the validity of the model under different operating conditions. In addition, the ability of the model to predict the rate of hydrogen production for other substrates, photocatalysts, and operating conditions was confirmed by comparing model predictions with experimental data from literature.