The forage-selection hypothesis (FSH) explains sexual segregation in ungulates as a function of different dietary requirements producing different levels of habitat optimality, whereas the reproductive-strategy hypothesis (RSH) explains sexual segregation as a function of different survival strategies between the sexes. Based on observations of habitat use by elk (Cervus elaphus) in Yellowstone National Park, Wyoming with regard to varying levels of wolf encounter risk, we found that our determination of whether the RSH or FSH best applied to sexual segregation varied by the scale at which we were measuring habitat use. At broad spatial scales we found no significant avoidance of wolves by elk. At the habitat scale we found that habitat use by elk was consistent with predictions of the reproductive strategy in that female elk used habitats that offered a balance of forage and escape terrain for themselves and calves, and that the degree to which escape terrain was present was dependent upon the risk of wolf encounter. At the scale of the habitat patch we found that differences in forage availability likely drove the differences in habitat use. Our results highlight the importance of scale when investigating habitat use, nonlethal predation effects, and sexual segregation in ungulates.