This study examines the impact of cover crop adoption on soil erosion levels in the United States (US) Midwest. Based on a novel county-level panel data set with information on soil erosion levels and remotely-sensed cover crop acreage, we estimate linear panel fixed effect econometric models and conduct a number of robustness checks to investigate the direct impact of cover crops on two major types of soil erosion (wind and water erosion). Although we find that counties with higher cover crop acreage have statistically lower soil erosion levels due to water, wind, or both, we believe that the magnitudes of the estimated effects are modest. Longer-term multi-year use of cover crops also do not seem to increase the soil erosion reducing effects of cover crops over time. Results from the empirical analysis provide further empirical evidence on the impact of cover crops on soil erosion based on data that captures farmer behavior at the county-level and covers a wider geographical region in the US. Our findings also give insights to policy makers in terms of further understanding the magnitude of the soil erosion benefits from cover crops.