Due to increasing human encroachment into the remaining habitat of many large carnivore species, there is an immediate need to understand the ecological and anthropogenic factors influencing carnivore space use decisions. In particular, knowledge of changes in space use in response to disturbance, and the costs associated with these changes, will be critical in guiding conservation efforts. To investigate the ecological factors influencing carnivore space use, we intensively radiotracked members of two large social groups (clans) of spotted hyaenas Crocuta crocuta in the Masai Mara National Reserve, Kenya. In addition, we studied the influence of livestock grazing by comparing space use between two study clans that differed dramatically in exposure to grazing. Logistic regression modeling indicated that space use in the absence of livestock was most influenced by the location of the clan's communal den. However, hyaenas were also found to select shrubland, areas of high prey density, and proximity to seasonal streams. Movements of hyaenas exposed to livestock grazing were most influenced by vegetation type, with a strong avoidance of open grass plains. Den location and prey density had less influence on space use decisions in the disturbed than the undisturbed clan. Livestock distribution did not directly influence hyaena movements either during daytime, when livestock were present, or at night. We suggest that direct livestock avoidance was unnecessary due to the observed increased use of vegetative cover by hyaenas exposed to grazing livestock. The greater distances from the den, and from areas of high prey density at which hyaenas were found in disturbed than undisturbed areas indicates potential energetic costs incurred by disturbed hyaenas. Our results therefore suggest that reduced vegetative cover, as is often found outside protected areas, may result in more dramatic modifications of hyaena movements in the presence of livestock. Keywords: carnivore, Crocuta crocuta, human disturbance, human-wildlife conflict, livestock, logistic regression, Masai Mara, spatial modeling.