Abstract: Chemical soil phosphorus (P) extraction has been widely used to characterize and understand changes in soil P fractions, however lacking adequately capture rhizosphere processes. Relying on two reduced P-input experimental stations in Taihu rice-wheat cropping rotation area, the biologically-based phosphorus (BBP) grading method was used to evaluate the availability and influencing factors of soil P during wheat-growing stage. The trial of long-term positioning P fertilizer reduction has been conducted for 7 years. The results show that, in Yixing Station, CaCl₂-P in soil treated with P fertilization at wheat season only (PW) had no significant difference with soil treated at both rice and wheat seasons (PR+W), while Citrate-P and HCl-P as well as Enzyme-P under PW treatments were significantly low (P < 0.05). As for Changshu Station, no significant differences were shown between three P-reducing treatments and PR+W; only Pzero treatment significantly reduced the concentration of soil HCl-P. Correlation coefficients between four BBP fractions and Olsen-P at two stations were different. Olsen-P was mainly from Citrate-P (R²=0.587, P < 0.01), HCl-P (R²=0.587, P < 0.01) and Enzyme-P (R²=0.531, P < 0.01) in Yixing, while mainly from HCl-P (R²=0.386, P < 0.05) and Citrate-P (R²=0.280, P < 0.05) in Changshu. The concentration of four BBP fractions followed the order of HCl-P > Citrate-P>Enzyme-P > CaCl₂-P. Furthermore, RDA showed that soil alkaline phosphatase (S-ALP) activity and pH were the dominant factors for the changes in soil P, and correlated with soil P fractions positively. These results may be helpful for understanding of the changes in soil P status while P-input was reducing.