Degradation of Fluoxetine antidepressant by Catalytic ozonation in aqueous medium was investigated using nano-gamma-alumina catalyst. Catalyst was synthesized via co-precipitation method and was characterized by XRD, FESEM, FTIR and BET Techniques. Controlled precipitation helped to successfully prepare nano-sized gamma-alumina particles using sodium carbonate as the precipitating agent and aluminum nitrate as the precursor. Artificial neural network (ANN) and central composite design (CCD) were used to model and optimize degradation of Fluoxetine and results of the two models were compared. Furthermore, impacts of the basic operational variable, i.e. inlet ozone concentration, initial Fluoxetine concentration, nano-gamma-alumina dosage and reaction time, were studied. Back-propagation (BP) learning for three-layer feed-forward ANN with topology 4:8:1 and trainscg algorithm was used for development of the ANN model. A considerable agreement was observed between the values predicted by the ANN and CCD models for removal of Fluoxetine and the experimental results. Findings declared superiority of ANNs in describing nonlinear behavior of the catalytic process and accuracy of the ANN model in predicting the efficiency values of Fluoxetine elimination. Pareto analysis demonstrated effectiveness of the all selected factors on efficiency of removal. Results showed that the most effective variable in catalytic ozonation of Fluoxetine is reaction time with 44.97% percentage effect. Maximum removal efficiency of 96.14% was obtained for 30 mg L-1 inlet ozone concentration, 1 g L-1 nano-gamma-alumina catalyst dosage, 30 min reaction time and 28.56 mg L-1 initial Fluoxetine concentration in optimum conditions.