Concentrations and annual loadings of molyhdate reactive P (MRP) and total (including particulate) P (TP) are reported from field drainage, catchment and erosion experiments in England. Annual losscs through field drains and in catchment runoff were 0.0374.74 kg MRPiha and 0.37-2.64kg TPiha, but those in surface runoff from cxperirnental plots measuring erosion wcre generally much greater (often > 3 kg MRPiha and up to 32 kgTPiha in a wet)ear). Amounts o f T P in drainflow and catchment runoff depended upon factors influencing soil dispersibility, such as particle size distribution and calcium carbonate content. The results to date suggest that P losses in surface runoff and erosion from arable fields to water are best limited by: (a) maximizing crop cover, using minimal cultivation practices and wherc possible planting crop rows across rather than up and down the slope, (b) avoiding cultivation practices that rcsult in dispersion of soil particles, and (c) avoiding appli-cation of P fertilizer to wet soils when rainfall is likely soon after application. Consideration should he given to maintaining field drains below peak efficiency to reduce subsurface P losses. I N T R O D U C T I O N s P inputs to surface waters are a key factor influencing A their trophic state, losses of even small amounts from soils to streams, ponds and lakes may pose a threat to water quality and aquatic wildlife. Previously it was thought that P losses from arable systems were negligible because the P applied in fertilizers is either taken up by crops or immobi-lized in soil by sorption on clay minerals or sesquioxides (Cooke, 1976). However, increasing eutrophication of many surface waterbodies in arable regions has been linked to increasing rates of soil erosion from fields cropped to winter cereals and accumulation of soil P through continuous