Phosphorus (P) loss in overland flow varies with spatial distribution of soil P, management, and hydrological pathways. The effect of flow time, flowpath length, and manure position on P loss in overland flow from two central Pennsylvania soils packed in boxes of varying length (0.5, 1.0, 1.5, 2.75, and 4.0 m long x 15 cm wide x 5 cm deep) were examined by collecting flow samples at 5-min intervals for 30 min (50 mm h(-1) rainfall) without and with 75 kg P ha(-1) applied as swine (Sus scrofa) manure over 0.5 m of the box slope length at distances of 0 to 3.5 m from the downslope collection point. Dissolved reactive P concentration was more closely related to the proportion of clay in sediment of overland flow before (r = 0.98) than after (r = 0.56) manure application. This was attributed to the transport of larger, low-density particles after applying manure. The concentration of dissolved and particulate P fractions decreased with increasing flowpath length, due to dilution rather than sorption of P by surface soil during overland flow. Total P loss (mainly as particulate P) from the Watson channery silt loam (fine-loamy, mixed, active, mesic Typic Fragiudult) was more than from Berks channery silt loam (loamy-skeletal, mixed, active, mesic Typic Dystrudept), even with manure applied. Thus, while P loss in overland flow is affected by where manure is applied relative to flowpath length, initial soil P concentration should not be discounted when looking at areas of potential P loss within a watershed.