Purpose <p>Optimization of phosphorus (P) fertilization from livestock manure can maximize crop production while protecting water quality. We assessed the effects of different forms (liquid, solid, and liquid composted with wheat straw) of swine manure application on yield and seed composition (N: P ratio) under corn-soybean rotation in a clay loam soil, Ontario, Canada.</p> Methods <p>Under an unified available nitrogen (N) rate (200&#xa0;kg N ha^− 1), P in each of liquid, solid and composted swine manures was applied to corn (<i>Zea mays</i> L.) in a 4-year corn-soybean (<i>Glycine max</i> L.) rotation at 0, 50 and 100&#xa0;kg P ha^− 1 (P-based), respectively, plus another P rate treatment from N-based application at 200&#xa0;kg available N ha^− 1, forming a series of four P rates. Chemical P fertilizer was also applied at 0, 50 and 100&#xa0;kg P ha^− 1, respectively, with the same unified N rate.</p> Results <p>The parabolic corn yield response to P rate was observed in liquid swine manure, but not in solid and composted swine manures. Corn seed N concentration and seed N: P ratio in alternative years were highly related to the yield, with optimum N: P ratio at 4.75. Soybean yields in both subsequent years responded to P rate of manure addition in a parabolic pattern with optimized P rate applied to preceding corn at 70&#xa0;kg P ha^− 1. Soybean yield with swine manure application was related to seed N: P ratio but varied from the year of crop rotation, suggesting the temporal changes in soil P availability with manure addition.</p> Conclusions <p>The relationship between corn yield and seed N: P ratio with manure application suggests that adjusting N: P ratio in manure might lead a way to optimizing corn yield and provide a new approach for manure application recommendation.</p>