TiO2 deposited on granular activated carbon (TiO2/GAC) was used for photocatalytic degradation of phenol. The effects of photocatalyst loading, initial substrate concentration and addition of an oxidizing agent as H2O2 were investigated using a one-factor-at-a-time experiment. Central composite design, an experimental design for response surface methodology (RSM), was used for the modelling and optimization of the phenol degradation. Analysis of variance (ANOVA) indicated that the proposed quadratic model was in agreement with the experimental case with R2 and R2(adj) correlation coefficients of 0.9760 and 0.9544, respectively. Accordingly, the optimum conditions for phenol degradation were a photocatalyst loading of two layers, initial phenol concentration of 34.44 mg L^(-1) and H2O2 concentration of 326.90 mg L^(-1). The TiO2/GAC was used for five cycles with phenol degradation efficiency still higher than 90%. Finally, the phenol that remained adsorbed on GAC was able to migrate to TiO2 and then photocatalytically be degraded.