Soybean seed isoflavones are affected by various factors, but their responses to phosphorus (P) fertilization is largely unknown. A field experiment was conducted at Jackson and Milan in Tennessee from 2008 to 2010 to investigate the effects of P application rates on concentrations and production of individual and total isoflavone in the form of aglycones in soybean seeds on low and high P soils under no-tillage. Five P application rates (0, 10, 20, 30, and 40 kg P ha−1) plus the University of Tennessee recommended P fertilizer rate based on initial soil-test P were arranged in a randomized complete block design with four replicates. Phosphorus application rates exerted more frequent and greater impacts on aglycone concentrations on the low testing P soil at Milan, particularly under the dry year of 2008. Application of 10–20 kg P ha−1 was adequate for producing high individual and total aglycone concentrations in seeds on the low testing P soil. Genistein concentration was more responsive to P fertilization than daidzein and glycitein concentrations. Individual aglycone concentrations were positively correlated (r = 0.25–0.91) among each other. Total aglycone was positively correlated with seed protein but not with oil in concentration. In conclusion, pre-plant fertilization of 10–20 kg P ha−1 is frequently beneficial for producing high concentrations of individual and total aglycones in soybean seeds on low P soils, particularly under dry years, without any significant reductions in other key quality attributes such as protein or oil.