High-intensity, short-duration (HISD) grazing has been hypothesized to simultaneously increase forage quality for livestock while maintaining habitat for grassland wildlife. I examined effects of HISD grazing on population size and perceived predation risk for rodents in an experimental coastal grassland in northwestern California. Mark-recapture data indicate that, relative to ungrazed areas, rodent populations plummeted on HISD-grazed fields. Depletable feeding trays were used to assess the giving up density (GUD) of food as a measure of predation risk. GUDs were consistently higher in grazed than in ungrazed fields, and they were positively correlated over time with the abundance of local raptors, suggesting that more abundant predators and less cover elevated perceived predation risk. However, raptors captured rodents less frequently over grazed fields, and the per capita risk of depredation by raptors was three times lower in grazed than ungrazed fields. Perceived predation risk may simply be a function of cover, regardless of true predation risk. Alternatively, the discrepancy in perceived and real predation risk as measured in this study may be reconciled by predation from other sources, such as meso-carnivores.