This paper presents the optimization process of a surface roughness model for the milling 1Cr18Ni9Ti. The model is developed in term of milling speed, feed per tooth and radial depth of cut. Therefore, the regression model predicting formula for surface roughness has been established by means of uniform design of experiment, and then the response surface methodology was applied to generate response contours of surface roughness. The experimental results indicate that the material removal rate can be improved by selecting optimal milling parameters without increasing the surface roughness. Moreover, it is seen that the feed rate is the most significant factor on the surface roughness.