A modified type of grassed waterway (GWW) with large hydrodynamic roughness has proven ability to reduce sediment load and surface runoff under conditions where best management practices on the delivering fields reduce sediment inputs that could otherwise damage the grass cover. It is unknown how such a GWW affects the loading of surface runoff with dissolved reactive phosphorus (DRP). The effect on DRP was tested in a landscape-scale study where DRP concentrations and loads in surface runoff were measured in two watersheds in which GWWs were newly installed and increased in effectiveness over time. Both watersheds were compared with paired watersheds without GWW installation; all watersheds were continuously monitored over 5 yr (1993-1997). Additionally, DRP concentrations were measured in open field and throughfall precipitation under growing grass and crops in field experiments, and DRP concentrations in surface runoff from straw covered surfaces were determined with laboratory rainfall simulation experiments. Dissolved reactive P in throughfall for the different cover types was highly variable, and the highest concentrations (up to 2.8 mg L(-1)) occurred especially during flowering of the respective crop and after frost events. Dissolved reactive P concentrations in runoff from straw-covered surfaces were slightly higher compared with those from bare soil. On average, there was a small difference in DRP concentrations between throughfall under growing crops and grass and in runoff from bare or straw covered soil surfaces. Hence, the introduction of a relatively small grassed area has little effect on the DRP concentration in surface runoff from the total watershed. This finding was supported by the watershed data, where watersheds with and without GWW showed similar DRP concentrations. No change in DRP concentrations occurred over the 5-yr period. Such GWWs will thus reduce the DRP load analogously to the reduction in total surface runoff.