Based on the background of mixed ownership reform of state-owned enterprises in China and the sample data of state-owned listed companies from 2004 to 2020, the impact of mixed ownership reform on total factor productivity of state-owned enterprises and its mechanism were empirically examined. Different from the treatment of dummy variables of mixed ownership reform in most literatures, the degree of mixed ownership reform (the shareholding ratio of the largest non-state-owned shareholder exceeds 5%) was quantified. In the test of the mechanism of action, commonly used mediating effect tests were used. Non-parametric Bootstrapping method was used to adjust the structural equation model of the estimated deviation to conduct multiple mediation effect test. This way could improve the testing power of the mediation effect and the interpretation of the results. The results show that mixed ownership reform has a significant positive effect on total factor productivity of state-owned enterprises, and the conclusion still holds after a series of robustness tests. Heterogeneity test shows that the positive effect of mixed ownership reform on total factor productivity of soes is more significant in local soes, the non-state-owned capital introduced as private capital, and the sample enterprises still holding state-owned shares after mixed ownership reform. The monetary compensation and equity compensation incentive of executives bear most of the mediation effect, especially the monetary compensation incentive of executives. After using the robustness test similar to benchmark regression, the mediation effect still holds significantly.