The objective of this study was to optimize the experimental conditions for maximum removal efficiency and minimum operating cost for removal of metribuzin pesticide from aqueous solution. Central Composite Design model was applied as Response Surface Design. In the model, independent parameters were ultraviolet irradiation time, persulfate anion dosage, TiO2 dosage, pH and response parameters were metribuzin removal (%) and operation cost. Analysis of variance (ANOVA) was used to describe the significance level of factors on the multiple performance characteristics considered. It has been found that the most suitable model is the quadratic model to explain the both responses and the correlation coefficient (R-2) of the relationship between the actual metribuzin removal and the predicted removal was calculated as 0.9518. As a result of multi-optimization, removal efficiency of 84.59 (%) and operation cost of 4.44 (euro/g) were obtained at ultraviolet irradiation time at 12.03 min, persulfate anion concentration at 4.909 g/L, TiO2 concentration at 0.197 g/L and pH at 4.