Surface application of animal manures has been shown to increase the concentrations of nutrients and trace elements in runoff. The objective of this study was to determine the effects of application of poultry litter on phosphorus (P) forms in runoff waters. Two poultry litters i.e. raw litter and granulated litter with urea were applied at 3 kg Water Extractable P (WEP) ha^-1, with WEP determined at three extraction ratios (1:10, 1:100, 1:200). When litters were applied at 1:10 WEP, the addition of total P from raw litter was 53 kg ha^-1 and 12 kg ha^-1 from granulated litter with urea. At 1:100 WEP, the addition of total P from raw and granulated litter with urea was 16 and 11 kg ha^-1. At 1:200 WEP, similar amounts of total P (11 kg ha^-1) were added from both litters. Rainfall simulations were conducted over a three-week period by packing wooden boxes (1-m long, 20-cm wide, 7.5-cm deep) with Captina silt loam (Fine-silty, siliceous, active, mesic Typic Fragiudult). A rainfall simulator was used to simulate a 7.0 cm h^-1 rainfall for 10 min as this was enough to generate continuous runoff from the boxes.  Runoff was collected, volumes recorded, and a composite sample was retained for analyses. For litters applied at the 1:10 and 1:100 WEP, dissolved reactive P (DRP) in runoff was greater from raw litter (13.8-24.8 mg L^-1) than granulated litter (9.2-10.0 mg L^-1). However, for litters applied at 1:200 WEP, the DRP was lower in raw litter (7.6 mg L^-1) compared with granulated litter treatment (9.9 mg L^-1). DRP concentrations continuously decreased from first to third rainfall event for all treatments. By third simulation event (21 day), the concentration of DRP were <0.4 mg L^-1 for all treatments except for raw litter (1:10), which was at 1.3 mg L^-1. This trend is attributed to the greater amount of P available in litters when litters were applied (1 day) to reduced amounts of P in the litters as time elapsed (7 and 21 days), which resulted in decreasing DRP concentrations in runoff at subsequent simulation events (7 and 21 days). The concentrations of dissolved unreactive P (DUP) and total dissolved P (TDP) followed a similar trend for all treatments as DRP. Interestingly, at third simulation event, DUP concentrations were greater than DRP for all treatments. This may be because of the flushes of organic P (i.e. DUP) due to wetting-drying cycles that proceeded before the third simulation event. These results suggest that surface application of raw litter can lead to greater P loss in runoff compared to granulated litter. While the magnitude of concentrations is much higher than the USEPA proposed maximum DRP concentration (1 mg L^-1) in surface runoff from agricultural soils, indicating that these concentrations may cause accelerated nutrient enrichment. However the concentrations reported in our study are from runoff boxes, where litter had minimum interaction with soil (because of physical separation of litter from soil) and secondly in watersheds, other soil biogeochemical processes may remove P from runoff by adsorption or dilution of concentrations may take place with less polluted water from other areas in the watershed.