【Objective】To investigate the variation of zinc (Zn) speciation in yellow paddy soil under different long-term fertilization and its effect on available Zn, a field experiment was designed in 1995. 【Method】There were six treatments: the control (no fertilizer, CK), P deficiency (NK), balanced application of chemical N, P and K fertilizer (NPK), single manure (M), lower organic and inorganic fertilizer combination treatments (0.5 MNP), and higher organic and inorganic fertilizer combination treatments (MNPK). The change process of soil available Zn and the temporal variation of different forms of Zn were studied. Also, the relationship between different forms of Zn and available Zn was analyzed with correlation and path analyses. 【Result】The results showed that the available Zn and Zn activation coefficient of soil with NK treatment decreased with the increase of experimental years. During this time, the linear fitting formula of the Zn activation coefficient reached a significant level, and the annual growth rate was -0.052 percentage points. Compared with NK treatment, the available Zn and Zn activation coefficient of soil with NPK treatment increased by 44.1% and 1.49 percentage points, respectively in 2019. Also, the contents of weak acid soluble Zn and reducible Zn increased by 15.9% and 5.3%, respectively. For treatments with organic fertilizer, the available Zn and Zn activation coefficient of soil increased linearly with the increase of experimental years, and the annual growth rate was 0.074-0.1244 mg·kg^–1and 0.032-0.063 percentage points, respectively. Compared with NPK treatment in 2019, the available Zn and Zn activation coefficient of soil in the manure treatments increased significantly by 6.3%-22.6% and 0.68-1.47 percentage points, respectively. For these treatments, the contents of weak acid-soluble, reducible, and oxidizable Zn increased significantly by 33.6%-84.0%, 32.5%-47.4%, and 25.5%-36.3%, respectively. In addition, the proportions of weak acid-soluble, reducible, and oxidizable Zn increased significantly by 0.6-1.3, 1.7-3.4, and 1.2-7.3 percentage points, respectively. On the contrary, the proportion of residual Zn was observed to decrease by 3.6-12.0 percentage points. Correlation analysis showed that soil available Zn was significantly positively correlated with weak acid-soluble, reducible, and oxidizable Zn, but significantly negatively correlated with residual Zn. Path analysis results showed that the direct path coefficient of weak acid-soluble Zn was the largest and had a significant direct positive effect on available Zn, residual Zn showed a significant direct negative effect on available Zn, while reducible and oxidizable Zn indirectly affected weak acid soluble Zn and had a strong indirect positive effect on available Zn. 【Conclusion】The weak acid-soluble Zn was the main source of available Zn in yellow paddy soil, while reducible and oxidizable Zn were the important cache database of available Zn. In agricultural production, the soil Zn efficiency could be improved by applying organic fertilizer reasonably to promote the transformation of Zn to weak acid-soluble, reducible, and oxidizable Zn, and avoid the accumulation of Zn in the residual state.