Phosphorus (P) management research in field crops has established that crops often utilize only a quarter to one third of inorganic fertilizer P while large amounts of residual P could accumulate in soil. Most of such studies have focused primarily on the plant available inorganic P, and only a few have identified the changes in the amount of individual inorganic P and organic P compounds. In the present 30-year field study in a semiarid subtropical irrigated soil, accumulation and transformations of residual fertilizer P in different P pools of soil, and its utilization, as induced by the effects of rate and frequency of P applications, in cereal and legume crop rotations were investigated. The study was initiated in 1975 on groundnut-based cropping systems where 4 levels of P (0, 20, 30 and 40 kg P₂O₅ ha^-1) were applied to: (a) summer-grown groundnut only (b) succeeding winter crop only and (c) both summer and winter crops for 25 years. First two frequencies represented alternative P treatment whereas the third one was cumulative P treatment. The application of fertilizer P was stopped in 2000 and the residual effect of P that had accumulated in different treatment plots was investigated for 5 years (2001-2005). Summer-grown groundnut showed small and inconsistent response to direct, residual and cumulative P application whereas winter-grown crops responded significantly and consistently suggesting 20, 40, and 60 kg P₂O₅ ha^-1 as adequate for groundnut, rapeseed-mustard, and wheat, respectively. Cumulative application of P produced yields comparable to those with direct applications. In no-P control plots, Olsen-P decreased slightly whereas organic P declined by about 50% after 25 years of cropping. In fertilizer-treated plots, organic P swelled up by 18 – 22 % and Olsen-P increased from initial 11.6 kg P ha^-1 to 14–34 kg P ha^-1 after 3 years and 16–58 kg P ha^-1 after 8 years, and further increase in between 8 and 25 years was negligible presumably due to reduction in the applied P rates after Year 7 and 17. Of the total fertilizer P added in 25 years, crops removed 21 to 54% whereas 33 to 64% accumulated in plough layer soil leaving 12 to 32% unaccounted for, which leached to lower soil layers. The residual fertilizer P was transformed into labile (9-19%), moderately-labile (12-22%) and non-labile (47-57%) pools of soil P. Residual P fertility was able to produce optimum groundnut and rapeseed yields for 3 and 5 years without applying fertilizer P in plots that had received alternate and cumulative fertilizer P applications in preceding 25 years, respectively. The results of this study demonstrated that (a) alternate application of fertilizer P to winter crops is required and P requirement of groundnut can be met from soil-derived and residual fertilizer P in the subtropical region, which helped enhance fertilizer use efficiency (42-54% as compared to 27-40% in cumulative P treatment), saved costly fertilizer (50%), and reduced excessive accumulation of residual P in soil (42-56%); (b) several fold increase in residual fertilizer P accumulation in soil and its accelerated conversion to non-labile P forms with long-term applications of fertilizer P to both crops each year without additional crop production have significant implications for soil testing and prediction of crop responses, leaching of excessive P in soil and consequent risk of contamination of groundwater; and (c) Olsen-P test alone is not adequate to reflect the depletion/ accumulation of bio-available P, and the status of different soil P pools in long-term fertilized soils must be determined to assess and make use of accumulated fertilizer P. Therefore, the best strategies for sustainable crop production and environmental safety are to produce optimum crop yields by (a) alternative P application only to winter crops, (b) periodic reduction in P rates according to accumulated P, and (c) occasionally without fertilizer P to enable the utilization of residual fertilizer P that had accumulated in the soil over the years leading to reduced leaching of P as well as substantial savings on this costly input by resource-limiting farmers.