The resistance spot welding process of galvanized steel sheet used in the body manufacturing of family car was studied, and the indexes of nugget geometry and tensile-shear strength of spot welds were tested. Four process parameters, namely welding current, electrode force, welding current duration and preheat current, and interactions among them were regarded as factors impacting indexes. Method using in mathematical models developing was nonlinear multiple orthogonal regression assembling design, which was optimized by the technology of variance analysis. The experimental results showed that more accurate prediction on nugget size and mechanical properties of spot welds can be obtained by the models optimized. With these prediction results, the optimization of welding process also was realized by the analysis to effect of the parameters and interactions on the welding quality.