【Objective】This study aimed to explore characteristics of the evolution of content and availability of soil phosphorus and their responses to phosphorus accumulation or loss (P balance) in reddish paddy soil under long-term fertilization varying in formulation under the double rice cropping system, in an attempt to provide theoretical bases for scientific application of phosphorus fertilizer in the double rice cropping areas of South China. 【Method】Based on a 35 -year long-term fertilization field experiment, researches were carried out on the evolution rules of soil total P, available P and soil P activation coefficient (PAC) in reddish paddy soil as affected by fertilization varying in formulation and history, calculations done of gain or loss of soil P each year and cumulative gain or loss of soil P in the long-term fertilization field experiment varying in formulation, and discussions made of relationships between soil total P, Olsen P, soil PAC and cumulative P balance. 【Result】Results show that soil total phosphorus, soil Olsen phosphorus and soil PAC in CK and Treatment NK (No phosphorus fertilizer applied) remained unchanged or displayed downward trends with the experiment going on. Soil total phosphorus in Treatment NK+PM (N and K fertilizer plus pig manure) exhibited a slow rising trend with the experiment going on, while soil Olsen phosphorus and PAC in the treatment did a downward trend. Soil total phosphorus in treatments NP, NPK, NP+RS and NPK+RS (chemical phosphorus fertilizer or plus rice straw) soared during the first ten years of the experiment, but the trend gradually leveled off during the 25 years that followed. Soil Olsen phosphorus in the above-listed-treatments rose sharply to high value in content during the first one to five years of the experiment, and then the rising trend began to level off with the years passing on. Soil PAC fell drastically in the treatments of no or low phosphorus input, but it ascended in the treatments applied with chemical phosphorus fertilizer or plus organic manure. Compared with the initial background value of the experiment field, soil PAC decreased by 33.2%, 29.7% and 16.6%, respectively in Treatments CK, NK and NK+PM, but increased by 66.2%, 60.6%, 65.6% and 52.9%, respectively in Treatments NP, NPK, NP+RS and NPK+RS. Treatment CK (No fertilization) led to soil phosphorus deficiency, while Treatment PM (applying only pig manure) nearly sustained P balance, and Treatments NP, NPK, NP+RS and NPK+RS resulted in apparent surplus of soil phosphorus. Phosphorus balance was very significantly related to soil total phosphorus, soil Olsen phosphorus and soil PAC. With an average surplus of 100 kg phosphorus per hectare, soil total phosphorus increased by 0.03 g kg^-1, and Olsen phosphorus by 1.20 mg kg^-1 and PAC by 0.09%. 【Conclusion】 Extraneous phosphorus inputs are the important factors that affect soil phosphorus and soil phosphorus availability. Long-term no phosphorus input and insufficient phosphorus input leads to soil phosphorus deficiency and hence reduction of soil phosphorus content and soil phosphorus availability. Applying chemical phosphorus fertilizer or chemical phosphorus fertilizer plus organic fertilizer promotes soil phosphorus accumulation and improves soil phosphorus fertility.