An experiment was conducted to evaluate the effects of four rates of straw mulch on runoff, infiltration, and erosion at a site in northeastern Oregon. Straw mulch densities of 0, 25, 50, 75, and 100 percent cover were tested using simulated rainfall applied to 3.33 m2plots located on a 16 percent north facing slope. Soil type is a Thatuna silt loam (fine-silty, mixed mesic Xeric Argialboll). The zero cover plots were replicated four times. The mean rainfall input rate of 28 mm/h did not vary significantly over the four replications. Median drop diameter of 1.3 mm at the center of the spray pattern is a nominal design characteristic of the rainfall simulator and is a representative size for natural rainfall in this area. Percent straw cover and normalized soil loss were highly correlated (r= 0.99), while no significant correlation between runoff volume or infiltration volume and soil loss was observed. This can be explained by the presence of a slowly permeable layer below the tillage depth and by the high organic matter content of this soil. The data suggest that smaller amounts of mulch (<25%) are of little value in erosion control and that a threshold value below which surface cover is ineffective may exist. This is a significant departure from the exponential relationship used to determine straw mulch effects on erosion. The data from this and earlier, similar experiments exhibit a large variability, especially in the erosion component, and emphasize a need for more experimental data to define the cover-soil loss relationship. There is almost no experimental field data for mulch covers less than about 30 percent, so further research is urgently needed to define the relationship for these lower cover conditions.