Long-term agricultural practices affect soil P status, and thus P bio-availability and vulnerability of losses. Previous studies mostly focused on short-term effects and limited to assessments within agricultural ecosystems. We evaluated the effects of long-term (≥45 years) consistent cropping systems (continuous corn: CC, continuous bluegrass: CB, and crop rotation of corn-oats-alfalfa-alfalfa: CR) and fertilization on changes in soil P status and losses in tile drainage discharge as compared to that in natural forest ecosystem in eastern Canada. Under the natural forest, soil P was predominated by stable P, followed by moderately stable inorganic P (Pi) and moderately labile organic P (Po). Agricultural production without fertilization reduced all forms of soil P, with labile and moderately labile P by 60-76% and moderately stable and stable P by 30-39%. Fertilization in agricultural systems reduced labile and moderately labile Po in the order of CC>CR>>CB, but increased labile and moderately Pi in the order of CB>>CC>CR. Fertilization increased P concentrations in tile drainage water by an average of 75-393%, depending on the P forms. Total soil P loss was the most in the CB plots, followed by CR and CC. regardless of fertilization. Total P loss was predominated by dissolved P in the fertilized CB plots, whereas, particulate P accounted for 62 to 74% of TP losses for CC, CR, and non-fertilized CB. Further analysis showed that the concentrations of dissolved Pi and PP, rather than the flow volume, predominated the contributions to the TP loss. This study stressed that the CC system significantly decreased P losses through reducing P concentrations in and flow volumes of tile drainage water. Continuous grass land may constitute an important source of P loading to fresh water resources in a long-term run.