American black (Ursus americanus) and grizzly bears (U. arctos) are sympatric throughout much of the grizzly bear's range, but information on how they share the landscape is lacking because distribution usually is not estimated simultaneously for both species. Here we analyze DNA data from noninvasively collected hair (using hair snags and rub trees) in a spatially explicit capture–recapture framework to study factors affecting the distribution of density for both black and grizzly bears in south Jasper National Park, Alberta, Canada, in 2014. Ninety-three black bears and 32 grizzly bears were detected. Black bears and grizzly bears showed different rates of detection for hair snags and rub trees, with hair snags being more effective in sampling both species. Female black bear density was greatest close to the town of Jasper, near roads, and in areas of higher amounts of closed canopy cover. Male black bear density was greatest at lower elevations with high canopy closure, and close to roads. Female grizzly bear density was greatest further from gravel roads, while male grizzly bear density was highest away from gravel roads and at medium canopy cover. We tested grizzly bears density as a predictor for black bear density and found that although habitat was the primary factor influencing black bear distribution, there was also a minor effect of grizzly bear density on male black bear density, and a slightly greater effect of grizzly bear density on female black bear density. The distribution of black bears within this study area puts them at a higher risk of conflict with humans. Our findings that grizzly bears prefer areas further from roads supports the use of habitat security thresholds to reduce human disturbance of grizzly bears.