Abstract: The continuous and excessive application of phosphorus (P) fertilizers and manure in intensive agricultural systems have led to soil P accumulation and progressive saturation of soil sorption capacity. This has significantly influenced soil P loss and P accumulation in aquatic ecosystems. The Taihu Lake Region of the Yangtse River Delta has for decades been a highly intensive agricultural production zone in China. Here, applications of chemical fertilizers and farmyard manure have been an effective method of improving soil fertility and productivity. Long-term applications of fertilizers and/or manure, often in excess of immediate plant uptake, have resulted in significant P accumulation and loss in this region. This has considerably increased the potential for eutrophication in the Taihu Lake. The change-point theory (with a soil Olsen-P content threshold above which the potential for significant P loss from soils to water systems occurred) has been considered to be scientific and useful in P management in agricultural soils. Up to date, however, there has been less report with respect to the Olsen-P change-point theory for paddy soils in the Taihu Lake Region. In this study, a long-term (13 years) P fertilization experiment in four P application doses (0 kg·hm^-2·a^-1, 30 kg·hm^-2·a^-1, 60 kg·hm^-2·a^-1 and 90 kg·hm ^-2·a^-1) was conducted in the Taihu Lake Region to evaluate the accumulation of Olsen-P under rice-wheat rotation cropping system. The experiment evaluated the environmental risks caused by P loss from soils to water systems. Topsoil (0~15 cm), surface water and leachates (30 cm and 60 cm) Olsen-P and total P (TP) contents were determined. Furthermore, a split-line regression model was used to estimate the risks of P loss from soils to water bodies and the change-point of soil Olsen-P in the Taihu Lake Region determined. Although, TP concentrations in surface water bodies and in 30 cm leachate significantly increased with increasing application rate of fertilizer-P, no significant increase was observed in the 60 cm leachate. As the risk of P runoff was mainly in the first 9 days after fertilizer-P application, it was advisable to control paddy field drainage during this period. Split-line regression analysis suggested that the change-points of Olsen-P content in surface soil triggering P leaching and runoff were respectively 26.0 mg·kg^-1, 24.8 mg·kg^-1. Fertilizer-P application at 60 kg·hm^-2·a^-1 and 90 kg·hm^-2·a^-1for 13 years resulted in respective soil Olsen-P accumulation of 26.9 mg·kg^-1 and 33.2 mg·kg^-1, which were all higher than the change-points. TP concentration of 30 cm leachate was also increased significantly. With continuous application of fertilizer-P at 30 kg·hm^-2·a^-1 for 13 years, soil Olsen-P content remained at (10.1±2.0) mg·kg^-1. This was sufficient for optimum rice/wheat growth without any risk of P loss. It was therefore not suitable to continuously apply 60 kg·hm^-2·a^-1 of fertilizer-P for a long time in paddy fields. The results demonstrated that intermittent fertilizer-P applications at 30 kg·hm^-2·a^-1 and 60 kg·hm^-2·a^-1 were suitable for rice-wheat rotation cropping system in the Taihu Lake Region.