Highway mortality of deer (Odocoileus sp.) is a nationwide concern. In 1991, 538,000 deer-vehicle collisions occurred nationwide. Property damage to vehicles, human injuries and fatalities, and potential impacts to local deer populations occur from deer-vehicle collisions. Techniques have been evaluated to reduce highway mortality of deer; however, an effective, cost-efficient solution does not exist for widespread use. If mitigative technologies are to be successful, we need to understand deer behavior and movement patterns associated with highway relationships. Most research about highway deer kills has focused on white-tailed deer (Odocoileus virginianus) in mixed hardwood habitat types. The following study pertains to mule deer (Odocoileus hemionus) in a mountain brush and sagebrush-grass zone. The study area was located at the site of the newly constructed Jordanelle Reservoir near Park City, Utah. Area roads were relocated due to inundation of existing highways. Preconstruction road-kill was docwnented to be 0.29 kills/km. Annual road-kill levels of 278 (5.9 kills/km) and 119 (2.5 kills/km) deer occurred along the new roads from October 1991 to October 1993. Even though there was a 64.2% reduction in observed deer density, second year mortality was still 9 times the pre-project kill. A study design of road-kill data collection and repetitive spotlight censuses was used to compare levels and composition of deer road-kills to that of the living population. Deer-vehicle collision levels tracked large population fluctuations. Deer behavior predisposed deer to mortality. Numbers of road-killed deer peaked in the fall of both years, coincident with breeding and hunting periods. Road-kill peaks also occurred in July and April of each year, respectively. Traffic characteristics, road alignment, and vegetative and topographic features were described relative to mule deer kill locations (recorded to the .10 mile). Traffic volume and percent vegetative cover were higher along US40 than either state route; road-kills were correspondingly higher along US40. Roads adjacent to agricultural areas along all routes sustained the fewest highway mortalities of deer. Deer approached roads along drainages; large drainages intersected highways in 79% of designated kill areas. Right-of-way vegetation and slope influenced kill locations.