The objective of this paper is to develop a Taguchi optimization method for low surface roughness in terms of process parameters when milling the mold surfaces of 7075-T6 aluminum material. Considering the process parameters of feed, cutting speed, axial-radial depth of cut, and machining tolerance, a series of milling experiments were performed to measure the roughness data. A regression analysis was applied to determine the fitness of data used in the Taguchi optimization method using milling experiments based on a full factorial design. Taguchi orthogonal arrays, signal-to-noise (S/N) ratio, and analysis of variance (ANOVA) are used to find the optimal levels and the effect of the process parameters on surface roughness. A confirmation experiment with the optimal levels of process parameters was carried out in order to demonstrate the effectiveness of the Taguchi method. It can be concluded that Taguchi method is very suitable in solving the surface quality problem of mold surfaces.