The high speed rotating arc welding sensor (RAWS) driven by motor with air axis is widely used in seam tracking system. Using the characters of the RAWS, the purpose of this study based on Genetic Algorithm (GA) is to optimize welding process parameters of the RAWS for obtaining satisfied seam geometry. The output variables are penetration, width and height of the seam geometry. These output variables are determined by input variables, which are the welding current, welding voltage, welding speed, rotating arc radius, rotating arc frequency and rotating arc direction. Experimentations are made according to the optimal welding process parameters, and the result shows that the GA is an effective method to optimize the welding process parameters of the RAWS. The described computational methodology enables obtaining a seam with desired geometry.