A factorial central composite orthogonal and rotatable design was employed to optimize the carbamazepine degradation using an integrated Fenton + Fenton-like oxidation process. The pH and the initial concentrations of hydrogen peroxide and ferrous and ferric ions were considered as the variables for the process optimization. A design of experiments procedure has been carried out in order to optimize the process as well as to study the interactions existing between the four variables under study. The initial concentration of hydrogen peroxide was found to be the most important variable conditioning the removal efficiency, followed by ferrous ion concentration, pH, and, finally, ferric ion concentration. The ANOVA test reported significance for 5 of the 14 involved variables. The response surface methodology technique was used to optimize carbamazepine degradation. Under optimal conditions (hydrogen peroxide concentration = 1.39 x 10(-4) mol L-1, ferrous ion concentration = 1.25 x 10(-5) mol L-1, ferric ion concentration = 1.68 x 10(-5) mol L-1, and pH = 3.52) total carbamazepine degradation was achieved.