Avoiding humans will be more difficult and energetically costly for animals as outdoor recreation increases and people venture farther into wildland areas that provide high‐quality habitat for wildlife. Restricting human access can be an attractive management tool to mitigate effects of human recreation activities on wildlife; however, the efficacy of such measures is rarely assessed. In 1982, Yellowstone National Park identified areas important to grizzly bears (Ursus arctos) to help protect critical grizzly bear habitat and reduce the likelihood of human injuries by bears. Referred to as bear management areas (BMAs), human access is restricted in these areas for 2–8 months each year, with timing and type of restrictions varying by area. We examined 2 datasets to evaluate grizzly bear selection of BMAs and differences of bear density in BMAs and non‐BMAs. First, we used 17 years of recent global positioning system telemetry data for grizzly bears to assess their selection of BMAs during periods when human access was allowed, and when access was restricted. We used step‐selection functions to test the hypothesis that bears spend time in places that allow them to avoid people and select quality food sources. There was support that grizzly bears differentially select for BMAs regardless of whether human access was restricted at the time, compared with areas outside BMAs, and that selection changed with sex and season. Only males during the summer and hyperphagic seasons changed their selection of BMAs based on whether access restrictions were in place, and overall, male bears preferred unrestricted BMAs (BMAs without restrictions in place). Females preferentially selected BMAs regardless of whether the area had access restrictions in place only during the mating season. Individuals varied widely in their preference for BMAs and access restrictions. Bears likely choose to spend time in BMAs based on available food resources rather than restrictions to human access. Supporting this interpretation, our analyses indicated that a greater proportion of BMA in an area was associated with higher densities of grizzly bear. Thus, restrictions to human access likely help reduce the potential for human–bear interactions, accomplishing one of the original objectives for establishing the BMAs. In 1982, Yellowstone National Park instituted seasonal (from 2 to 8 months) restrictions to human access in areas thought to be important to grizzly bears (Bear Management Areas, BMAs), to help protect critical habitat for grizzly bears and reduce the likelihood of human injuries by bears. We evaluated selection of these areas by grizzly bears and found support that they selected BMAs, but that bears likely chose to spend time in these areas for the resources available rather than the restrictions to human access. We also found higher densities of grizzly bears inside BMAs, compared with outside these areas, suggesting that restrictions of human access likely helped to reduce the potential for human‐bear interactions.