Objective In order to monitor the dynamic change of memory T cell subsets in mouse peripheral blood, an optimal protocol with combination of minimally invasive blood sample technique and flow cytometry analysis was established. Methods The blood was sample via saphenous vein, and the four color compensation matrix of flow cytometry was established with fluorescence compensation beads, then the ratios of naive T cells, central memory T cells and effector memory T cells among blood samples were analyzed using BD Calibur equipped with two laser. The following aspects were investigated to optimize the flow cytometry protocol:(1) blood sampling volume; (2) centrifugation of blood or not; (3) the concentration of detecting antibodies; (4) wash after the lysis of erythrocyte or not. Results (1) Ten to fifty microliter of blood could be sampled via saphenous vein without animal anesthesia, which would satisfy the requirements of serial blood sampling and analysis, ten microliter blood could be fulfill the requirements of multi-color flow cytometry analysis and decrease the demand of antibodies. (2) Demand of antibodies could be used to detect decreased blood after high speed centrifugation and removal of serum or plasma. (3) Optimization of antibody concentration would decrease the volume of antibodies and potential interference of the background fluorescence without influencing the accuracy and reproducibility. (4) Wash after the lysis of erythrocyte could furthermore decrease the background fluorescence of samples, but increase the operation time, which could be analyzed without wash immediately. Conclusion Combination of blood sampling via saphenous vein and optimal flow cytometry analysis protocol could help to monitor the change of memory T cell subsets in vivo of mouse in a serial of time-points.