Understanding relationships between soil P and P loss with surface runoff is important to assess the risk of P delivery from fields to surface water. This Iowa study was part of a cooperative effort to characterize these relationships for important USA agricultural soils. Over 400 simulated rainfall events were conducted on recently tilled corn and soybean fields representing the soil series Clarion (Typic Hapludoll), Kenyon (Typic Hapludoll), and Tama (Typic Argiudoll). Soil samples from the plot area (1.5 by 2.0 m) were analyzed for P with five routine and environmental tests. The STP ranges across plots (mg P kg^-1, 5 cm depth) were 4-732 for Bray-1 (BP), 5-1012 for Mehlich-3, 2-511 for Olsen (OP), 3-393 for Fe-oxide impregnated paper, and 1-245 for water-extractable tests. Rainfall was applied at 7.7 cm h^-1 and runoff was collected for 30 min. Across all simulations mean runoff concentrations for bioavailable P (BAP) were greater than those for dissolved reactive P (DRP) by a factor of 2.7 for Tama, 1.7 for Kenyon and 3.2 for Clarion soil. Runoff DRP and BAP increased linearly with increasing STP as measured by all tests. Correlation coefficients between STP and DRP or BAP ranged from 0.86 to 0.93 across soils and soil tests. The ranking of soils for DRP and BAP rates of increase with increasing STP was consistent across tests, although relative differences varied. For example, slopes for Clarion, Kenyon, and Tama were 0.0018, 0.0017, 0.0015 mg DRP L^-1per mg BP kg^-1, respectively, but were 0.0039, 0.0040, and 0.0021 mg DRP L^-1 per mg OP kg^-1. A smaller slope for Tama soil was consistent with soil properties suggesting higher P sorption capacity than for other soils.