Comprehensive understanding of the fate of soil P as related to various management practices is essential to develop nutrient management plans that sustain crop production while minimizing adverse impacts on water quality. A long-term study was conducted on a Brandon loam soil from 1992 to 2003 to evaluate changes of soil P fractions as a function of cropping system and fertilization practices. Treatments included factorial combinations of three cropping systems [continuous corn (CC), corn following alfalfa (CA), and corn following soybean (CS)] and three fertilization practices [inorganic nitrogen (N) at zero (ZN) and 100 kg ha-1 (100N) both with fertilizer P added based on soil test, and dairy manure at 12.5 Mg ha-1 (DM)]. Soils were sampled in the fall of 2003 and soil P was fractionated using the modified Hedley's sequential extraction procedure. All soil P fractions, except for residual P (Res-P), were not affected by cropping system and N application. Addition of DM increased water extractable inorganic P (Pi) by 7.6 mg kg-1, NaHCO3 extractable Pi by 43.2 mg kg-1 and Po by 11.7 mg kg-1, and HCl extractable P by 67 mg kg-1, relative to the ZN plots. However, addition of DM decreased NaOH extractable Pi by 10.4 mg kg-1 and Po by 38.2 mg kg-1, compared with ZN plots. Addition of DM combined with cropping systems including a legume crop (CS and CA) increased soil Res-P, while addition of DM with CC decreased soil Res-P. The increases of both labile P (water-Pi) and stable P (HCl-P and Res-P) suggest that long-term addition of DM can have both immediate and long-term impacts on water quality and crop growth, which need to be taken into consideration when planning nutrient management programs.