Our work aimed to propose a manual method of counting CD4 T lymphocytes which is an alternative magnetic immunoseparation followed by a reading with a fluorescence microscope as an alternative to the automated flow cytometry. This alternative technique is easier for use, less expensive and could answer the difficulties encountered for the monitoring CD4 T cells count in developing countries. The specific objectives were: 1) to train the technicians of the peripheral sites in order to make the numeration of the CD4 T lymphocytes more accessible at the peripheral level; 2) to equip the sites with necessary facilities for the T lymphocytes CD4 count; 3) to put in place a system of quality control permitting the reliability of the results. A hundred and fifty patients have been enrolled in three care services for people living with HIV/AIDS in Dakar. This population was constituted of 119 seropositive and 31 seronegative patients acting as control group to have some patients with high rates of T lymphocytes CD4. For the follow-up at peripheral level, the patients were constituted of the active line of the patients living with HIV/AIDS supported in the targeted sites. The measurements allowed studying concordances for different rates of lymphocytes: 0 to 199, 200 to 499 and over 500 cells by mm3. The results showed also a very good correlation (r = 0.97 or r = 0.98 according to the operator) between the two methods for CD4 rates inferior to 500 cells by mm3 among both the negative group and the HIV positive patients. We also discussed the profit of decentralization for the program and the patient, as well as the setting up of an external quality control to validate the alternative technique. According to the results, the Dynabeads is well correlated with the Facscount. It is a technique that can be used as an alternative in the zones with limited resources, low prevalence and for a small number of samples.