Phosphorus (P) in irrigation runoff may enrich offsite water bodies and streams and be influenced by irrigation water quality and antecedent soil surface conditions. Runoff, soil loss, and P fractions in runoff using reverse osmosis (RO) water or mixed RO and well water (RO/ Tap) were studied in a laboratory sprinkler study to evaluate water source effects on P transport. A top- or subsoil Portneuf silt loam (coarse-silty, mixed, superactive, mesic Durinodic Xeric Haplocalcid), either amended or not amended with manure and/or with cheese whey, with Olsen P from 20 to 141 mg kg(-1) and lime from 108 to 243 g kg(-1), was placed in 1.5 x 1.2 x 0.2-m-deep containers with 2.4% slope and irrigated three times from a 3-m height for 15 min, applying 20 mm of water. The first irrigation was on a dry loose surface, the second on a wet surface, and the third on a dry crusted surface. Surface (ca. 2 cm) soil samples, prior to the first irrigation, were analyzed for Olsen P, water-soluble P (Pws), and iron-oxide impregnated paper-extractable P (FeO-P) analyses. Following each irrigation we determined runoff, sediment, dissolved reactive phosphorus (DRP) in a 0.45-microm filtered sample, and FeO-P and total P in unfiltered samples. Soil surface conditions had no effect on P runoff relationships. Water source had no significant effect on the relationship between DRP or FeO-P runoff and soil test P, except for DRP in RO runoff versus water-soluble soil P (r2 = 0.90). Total P in RO runoff versus soil P were not related; but weakly correlated for RO/Tap (r2 < 0.50). Water source and soil surface conditions had little or no effect on P runoff from this calcareous soil.