Legumes have been shown to enhance bioavailability of phosphorus (P) from sparingly soluble pools, yet this functional trait remains underutilized in agriculture, and is untested at decadal scales. Management and legume presence effects on temporal soil properties were evaluated in a 17-year field crop experiment using soil samples collected in 1992, 2000 and 2006. Management systems compared included: (1) conventional corn–soybean–wheat rotation (C–S–W), (2) organic (C–S–W+red clover), (3) alfalfa and (4) early successional field. To evaluate the effects of long-term management versus recent management (residues and P fertilizer) on P and bio-availability to soybean, subplots of soybean were established with and without P-fertilizer (30 kg P ha−1), and compared to subplots and main plot with the long-term system. We evaluated soil properties (C, total P, Bray extractable inorganic P, particulate organic matter phosphorus) and soybean P uptake, biomass and yield. Recent fertilizer P inputs had no detectable influence on soil P, and total soil P stayed stable at ~350 mg P kg−1, whereas inorganic P (Pi) declined from an initial value of 54 to an average of 35 mg P kg−1. A P balance was constructed and showed a net loss of −96.7 kg P ha−1 yr−1 for the organic system, yet Bray-Pi and soybean P uptake were maintained under organic production at similar levels to the conventional, fertilized system. Particulate organic matter P was 57, 82 and 128% higher in organic, alfalfa and successional treatments, respectively, compared to conventional. A similar pattern was observed for soil C, soybean yield and bioavailable P, which were 20–50% higher in the organic, alfalfa and successional systems relative to conventional. This study provides evidence that long-term management history influences bioavailability of P.