Faecal indicator organisms (FIOs) are major pollutants in many catchments world-wide, with streamside pastures on livestock farms being potentially significant sources. Hitherto, few empirical studies have quantified FIO fluxes from such areas or investigated streambank fencing (SBF) and other possible mitigation measures. The aim of this two-phase (before/after intervention) study of the effectiveness of SBF was to generate an empirical evidence-base to enable regulatory authorities to make better-informed decisions concerning the implementation of this measure. It was undertaken during the summer bathing season along a 271 m stream reach in the River Tamar catchment, SW England. The study included: cattle distribution surveys; monitoring of changes in E. coli (EC) and intestinal enterococci (IE) concentrations and fluxes down the reach and of concentrations in ditch flow and surface runoff; phage tracer studies of surface runoff from pasture land; and experimental streambed trampling to investigate streambed FIO sources. The results show that cattle spend a disproportionately large amount of time in the watercourse/riparian zone along unfenced streams; identify direct defecation to the stream by wading livestock and the release/mobilisation of FIOs from cowpats by surface runoff from the adjacent pastures at times of high flow as key transmission routes; and demonstrate that FIOs become incorporated within streambed sediments, from which they may subsequently be released by trampling. Partial exclusion of cattle through SBF with a drinking bay greatly reduces the time cattle spend in streams. Total exclusion SBF, with provision of an alternative drinking supply, considerably reduces FIO load inputs to the stream reach, e.g. at times of high flow, which are critical in terms of pollutant fluxes to coastal waters, the mean EC and IE input loads to the reach fell by 0.842 and 2.206 log10, respectively.