In order to ensure high quality of micro scale laser shock peening (μLSP), an appropriate analysis model is established for the numerical simulation of the process, and then some key technical issues in the simulation are studied and solved. By extracting the characteristics of residual stress field and comparing the calculated data with the process parameters, the critical parameters affecting the value of the characteristics of residual stress field induced by μLSP are searched out. The Box-Behnken optimization program based on the response surface methodology (RSM) is developed to search out the influence of the identified critical parameters and the interaction of the parameters to characteristics of the residual stress field. A quadratic regression model for predicting the average surface residual stress which is one of the characteristics of residual stress field is established, the optimal setting for each process parameter is predicted also.