Abstract: Objective To discuss the realization of linear mixed model in SPSS and SAS for repeated measurements with missing values. Methods Data from a depression therapy with missing values were analyzed using GLM and MIXED procedure, so as to evaluate the therapy effect. The measurements are depression levels after treatment. Results When missing values exist, analysis with GLM procedure will exclude the subjects from further calculation, while analysis with MIXED procedure will include the subjects in further calculation. They lead to different statistical results. Conclusion MIXED extends repeated measures models in GLM to allow an unequal number of repetitions and it will still be efficient under more complex situations where data units are nested in a hierarchy. SPSS and SAS both can get efficient estimators for either balanced or unbalanced data.