This study aims to discuss the effect of preventing pathological changes and cognitive degeneration of Tg2576 mice by inoculating the subunit fragment of Aβ vaccine. Thirty-two Tg2576 mice were randomly divided into four groups, each having eight mice: Group I, the control group, inoculated with adjuvants; Group II, the Aβ42 group, inoculated with Aβ42 vaccine; Group III, the Aβ1―15 group, inoculated with Aβ1―15 vaccine; and Group IV, the Aβ36―42 group, inoculated with Aβ36―42 vaccine. The titer of the serum anti-body against Aβ42 (Group II) was significantly higher than that of the control group (Group I), and a low level of antibodies could be detected in the brain homogenate in the three vaccine-inoculated groups. Morris water maze test showed that the Aβ42 group, Aβ1―15 group and Aβ36―42 group were obviously im-proved compared with the control group. The cultured splenocytes sampled from each group were induced by Con A or their respective antigens, and the cell proliferation of the three vaccine-inoculated groups was significantly higher than that of the control group. In the Aβ42 group, IL2 and IFN-γ were relatively low and IL4 and IL10 were relatively high. By contrast, IL4 and IL10 were much higher in the Aβ1―15 group and IL2 and IFN-γ were much higher in the Aβ36―42 group. The immunohistochemical test showed a large number of senile plaques in the brain cortex and hippocampus of the mice in the con-trol group, no senile plaque in the brain of the Aβ1―15 group and Aβ42 group mice, and a small number of senile plaques in the brain of the Aβ36―42 group mice. The results suggest that the subunit fragment of Aβ1―15 vaccine could prevent not only cognitive and behavioral degeneration but also Aβ deposition and formation of senile plaques in Tg2576 mice. This study aims to discuss the effect of preventing pathological changes and cognitive degeneration of Tg2576 mice by inoculating the subunit fragment of Aβ vaccine. Thirty-two Tg2576 mice were differentiated into four groups, each having eight mice: Group I, the control group, inoculated with adjuvants; Group II, the Aβ42 group, inoculated with Aβ42 vaccine; Group III, the Aβ1-15 group, inoculated with Aβ1-15 vaccine; and Group IV, the Aβ36-42 group, inoculated with Aβ36-42 vaccine. titer of the serum anti-body against Aβ42 (Group II) was significantly higher than that of the control group (Group I), and a low level of antibodies could be detected in the brain homogenate in the three vaccine-inoculated groups. Morris water maze test showed that the Aβ42 group, Aβ1-15 group and Aβ36-42 group were significantly im-proved compared with the control group. The cultured splenocytes sampled from each group were induced by Con A or their corresponding antigens, and the c ell proliferation of the three vaccine-inoculated groups was significantly higher than that of the control group. In the Aβ42 group, IL2 and IFN-γ were relatively low and IL4 and IL10 were relatively high. By contrast, IL4 and IL10 were much higher in the Aβ1-15 group and IL2 and IFN-γ were much higher in the Aβ36-42 group. The immunohistochemical test showed a large number of senile plaques in the brain cortex and hippocampus of the mice in the con-trol group, no senile plaque in the brain of the Aβ1-15 group and Aβ42 group mice, and a small number of senile plaques in the brain of the Aβ36-42 group mice. The results suggest that the subunit fragment of Aβ1-15 vaccine could prevent not only cognitive and behavioral degeneration but also Aβ deposition and formation of senile plaques in Tg2576 mice.