Based on the variable selection methods and robust estimation methods presnted in literatures,a robust variable selection method was proposed for longitudinal data,and the robustness of the proposed method was evaluated by simulation study.The proposed method was then applied in the analysis of a real data set.It is found from the simulation and analysis of the real data that the proposed robust variable selection method can estimate correctly the regression parameters and select properly the important covariant variables,and it is also robust against outliers in the data set.