Methods The Tai Chi group（ n=94） and the control group（ n=61） were assessed at baseline for participants' cognitive function with the measures of Mini-Mental State Examination（ MMSE）， Montreal Cognitive Assessment（ MoCA）， verbal fluency， digit symbol， copy， trail making test， picture description， repetition， associative learning， episodic memory， similarity and calculation. Four years later， a follow-up assessment was conducted. We assessed the cognitive function of the participants in each group to conduct inter-person comparisons before and after 4 years， and to compare the differences of cognitive function change between the two groups. Results At baseline， the only significant difference between the two groups was the total score of the baseline MMSE［ （28.21±2.04） vs.（ 27.09±2.17），t=2.485， P=0.014］. Compared with the baseline， the cognitive function of the control group declined in most dimensions after 4 years（ P＜0.05） except the repetition score. Similarly， there was no significant decline in the repetition score of Tai Chi group. We used each group's score change before and after 4 years in different cognitive function dimensions to compare the cognitive function decline in the two groups. We found that， compared to the control group， there was a narrower score change in the Tai Chi group in terms of MMSE［ 1（0，2） vs. 2（0，3），Z=2.719，P=0.007］， MoCA［ （1 0，4） vs. 4（3，6），Z=-5.697，P＜ 0.001］， trail making test［ 6（0，13） vs. 11（2，24），Z=-2.359，P=0.018］， episodic memory［ 0.5（0，1.5） vs. 1（0.5，2），Z=-2.292，P=0.022］ and similarity［ 1（0，2） vs. 2（0，4），Z=-2.678，P=0.007］. Conclusions Tai Chi has a positive effect on slowing cognitive function decline in senior individuals. Longterm Tai Chi exercise may be beneficial to the retention of cognitive function in the senior population.