Response surface methodology (RSM) was used for the optimization of the baking process of cookies made with onion powders (onion cookie). Experiments were carried out according to a central composite design, selecting onion powder content $(1.15{sim}8.86%)$ and baking temperature $(155.9{sim}184.1^{circ}C)$ as independent variables and pH, titratable acidity, moisture content, density, spread factor, CIE color parameters $(L^{ast}-$, $a^{ast}-$, and $b^{ast}-values)$, and hardness as response variables. pH decreased with increasing amount of onion powder and baking temperature. Moisture content also decreased as the baking temperature increased. $L^{ast}-$ and $b^{ast}-values$ decreased but $a^{ast}-values$ increased with increasing onion content and baking temperature. Most polynomial models developed by RSM were highly effective to describe the relationships between the studied factors and the responses. Overall optimization, conducted by overlaying the contour plots under investigation, was able to point out the optimal range of the independent variables within which the six responses were simultaneously optimized. The point chosen as representative of this optimal region corresponded to 4.01% onion powder content and $161.84^{circ}C$ baking temperature. The predicted dependent or response values in the optimal region were: pH＝6.87, moisture content＝2.77%, $L^{ast}-value＝68.45$, $a^{ast}-value＝1.98$, $b^{ast}-value＝34.64$, and spread factor＝9.41.