This project examined the application of grass buffer strips to improve runoff water quality from irrigated pastures in the Sierra Nevada foothills. These flood irrigated pastures range up to 30% slope, and can generate significant runoff. Three experiments were conducted to determine: 1) the partitioning of nitrogen (N) between soil, plants and runoff within buffers; 2) whether buffer capacity for N decreases over time as buffer vegetation matures in the absence of grazing/cutting; and 3) the efficiency of buffers to attenuate E. coli, total phosphorus (P), dissolves organic carbon (DOC), and suspended solids in a rotationally grazed pasture scenario designed to offset the timing of grazing bouts from irrigation events. These experiments were conducted on irrigated pasture – buffer runoff plots at the UC Sierra Foothill Research and Extension Center near Brown’s Valley, CA. Buffer size treatments were 0, 8, and 16 m and grazing – irrigation offset treatments were 2, 15, and 30 days. We used the nitrogen isotope (15N) method in Experiments 1 and 2. Vegetative uptake was a major mechanism for attenuating new N in irrigated pasture systems, and nutrient cycling within vegetative buffers was serving as both a sink and a source for N in runoff. Buffers were effective for attenuating nitrate (NO3-15N), slightly more effective for ammonium (NH4-15N), and least effective for dissolved organic nitrogen (DON-15N). For DON, the 16 m buffer was actually less effective than the 8 m buffer, indicating that the 16 m buffers themselves were serving as a source for this less plant-available form of N. Monthly cutting of buffer vegetation doubled 15N uptake compared to uncut buffers, confirming that regular cutting and harvest of buffer vegetation increases vegetative buffer efficacy for N uptake. Under the irrigation application – runoff – transport capacity scenario examined in this study, we could attribute no significant reduction in dissolved organic carbon, total suspended sediment, E. coli, or total phosphorus load (kg/ha) in irrigation runoff to 3 year nongrazed/ cut vegetative buffers either 8 or 16 m in width. DOC load was actually significantly (P<0.05) increased on plots with a 16 m buffer, and there were apparent increases in load for TSS, VTSS, and E. coli for both 8 and 16 m buffer widths compared to no buffer control plots. Pollutant load was positively related to runoff volume, indicating that reductions in runoff volume will result in reduced pollutant transport. Pollutant load was significantly reduced by increasing days rest from grazing prior to irrigation from 2 to 15 days. Extending this rest to 30 days gained only slight additional reduction in pollutant load. The general failure of buffers to reduce DOC, TSS, VTSS, E. coli, and P loads in Experiment 3 under the high irrigation application – runoff – transport capacity scenario examined in this study should not be extrapolated to conclude that vegetative buffers have no merit for water quality improvement in this system. Rather, it is clear that application of buffers to irrigated pastures without a simultaneous effort to balance irrigation rates with soil infiltration capacity and plant-soil water demand will certainly not achieve water quality protection. It is also clear that management of buffer vegetation will be required to maintain buffer capacity for nutrients, and to reduce the potential for buffers to become a source for DOC and DON, and habitat for rodents shedding E. coli in their feces.