A field experiment was conducted in 2011 to study the effects of phosphorus (P) fertilizer application on the phosphorus fractions and microbial diversity of rice - rape crop rotation. Two treatments, NK (-P) and NPK (+P), were selected for the experiment. The soil layer was divided into 0~10、10~20、20~30 and 30~40 cm for sampling and analysis. Soil available P, total P and organic P were determined by conventional analytical methods. Chemical continuous extraction method was used to determine the inorganic P fractions in the soil samples, and Illumina-miseqpe250 platform was used for high-throughput sequencing of surface soil (0~10 cm) microorganisms. The results showed that, compared with the CK (-P treatment), long-term application of P fertilizer (+P) could significantly increase the content of total P, inorganic P and available P in each soil layer, but the organic P content did not increase significantly. From the perspective of available P content in soil layer, P content in soil layer of 0~10 cm and 10~20 cm could be significantly increased by application of P fertilizer, but the effect on the bottom layer of 20~30 cm and 30~40 cm was not obvious. Compared with -P treatment, long-term application of P fertilizer reduced the Alpha of microbial community diversity and changed the structure of microbial community. At the level of phylum, the dominant bacteria were Chloroflexi, Acidobacteria and Proteobacteria. The relative abundance of Chloroflexi and Acidobacteria were increased, while that of Proteobacteria decreased. Ascomycota and Basidiomycota were the dominant phylum, the relative abundance of Ascomycota decreased and that of Basidiomycota increased. Compared with CK, although the application of P fertilizer caused a significant decrease in the number of bacteria and fungi in the layer soil in each taxa, the relative abundance of the dominant flora also decreased from the perspective of the genus level. The results of redundancy analysis showed that the content of Al-P and Fe-P was closely related to the abundance of various bacteria and fungi. It could be seen that long-term phosphorus deficiency would induce an increase in the number of microorganisms to activate soil nutrients, and a sufficient supply of phosphorus fertilizer would meet crop nutrient requirements and maintain the dynamic balance of soil microbial communities. Therefore, it is necessary to optimize the application of P fertilizer in the entire crop rotation system to promote the activation and release of soil phosphorus, thereby reducing the amount of P fertilizer and improving its recovery rate.