BACKGROUND: Heterogeneous photocatalysis is influenced by a number of parameters involving synergistic effects; hence, an experimental strategy design that considers interactions between the main variables is needed. The response surface methodology was applied for the investigation of photodegradation of 20 mg L-1 Orange II in aqueous solutions and for optimization of color removal efficiency. Preliminary studies were performed to identify the parameters to be selected for optimization. RESULTS: The input variables considered for experimental design were: solution initial pH, oxidizing agent (H2O2) initial concentration and UV-A irradiation time. The multivariate experimental design allowed the development of a quadratic response surface model to be used for the prediction of color removal efficiency over the full range of the experimental region. Under the optimum conditions established in the region of experimentation (pH = 6.9, [H2O2](0) = 183 mg L-1 and t = 32 min), a 100% color removal efficiency was obtained in experiments. CONCLUSIONS: It was found that the variables considered have important effects on color removal efficiency. The results demonstrate that the use of experimental design strategy is indispensable for successful investigation and adequate modeling of the process because the interdependence of the reaction parameters cannot be neglected. (C) 2008 Society of Chemical Industry