The purpose of this study was to identify road and habitat characteristics associated with black bear-vehicle collisions in Yosemite National Park and to suggest proper mitigations to reduce their occurrence. Black bear-vehicle collision data collected by Yosemite National Park staff between 1995 and 2011 were used to identify variables associated with collisions. Geographic Information System (GIS) mapping software was used to map and split Yosemite roads into 1 km segments. After measuring road and bear habitat-related variables along each road segment, logistic regression analyses showed that segments with collisions were associated with crossing sites, understory vegetation, curves, close proximity to meadows, and a flat outbound shoulder slope. GIS spatial pattern and hot spot analysis were then used to group segments by their relative frequency of collisions: zero, low, moderate, and high. Logistic regression analyses of those same road segments, now grouped by their collision frequency, showed that segments with high frequencies of collisions were associated with a lack of visibility, fewer crossing sites, high understory cover, steep shoulder slopes, and close proximity to human development and meadows. The findings of this study were used to suggest effective and appropriate mitigation strategies for reducing collisions between bears and vehicles.