Highways fragment important habitat and can serve as impassable barriers to wildlife. Installing wildlife crossing systems creates opportunities for safe animal movement, improved human safety, and reduced property damage. Research identifying key zones of connectivity can aid in providing recommendations for locations of these systems and other tools to improve highway safety. Although elk-vehicle collisions have been documented across northwestern California, the factors associated with these highway crossing zones have not been studied in this region. To address this gap in research, the objectives of this study were to determine the frequency of elk highway crossings on hourly, daily, and seasonal time scales, and to evaluate how various environmental and anthropogenic factors influence the probability of elk crossing US-101 in northwestern California. Using GPS collar data collected between December 2016 and February 2024 from 45 female Roosevelt and tule elk from 16 established herds adjacent to US-101, a resource selection function was used to predict the probability of elk highway crossing frequency across the study area. Environmental and anthropogenic covariates used in the model include: distance to forest edge, forest cover, land cover type, Terrain Ruggedness Index, slope, distance to urban area, distance to secondary road, number of highway lanes, presence of underpass, traffic volume, season, and time of day. The openness ratio of highway undercrossings within established elk home ranges was calculated and elk use of underpasses was determined through the deployment of trail cameras from June to October 2024. These results will contribute to future management decisions regarding wildlife crossing systems and increasing connectivity to mitigate the negative impacts of highways on elk and other large mammals as well as reduce conflict and safety risks with humans.