Roads negatively affect wildlife populations through the direct impacts of wildlife- vehicle collisions and the indirect impacts of restricted movement and altered habitat use. These impacts can scale up to loss of connectivity among populations, leading to local extinctions. South Texas is home to the last remaining breeding populations of ocelots in the United States, and these endangered felids are vulnerable to the negative effects of roads. Land conversion and road construction have reduced and fragmented viable habitat. As a result, vehicle collisions are the primary source of known ocelot mortality, and lack of connectivity between populations has led to inbreeding and genetic isolation. Road mitigation, including crossing structures and fencing, is a necessary component of population recovery efforts. To support mitigation, we evaluated landscape and road attributes associated with elevated road mortality risk across the south Texas road network using a presence-only species distribution modeling approach. We also assessed whether a sympatric carnivore, the bobcat, could serve as a surrogate to inform mortality risk. Areas where core ocelot habitat intersected roadways represented the greatest mortality risk, reflecting ocelot reliance on intact habitat and avoidance of fragmented landscapes. Bobcat road mortality risk had similar relationships to landscape features, with nearly all areas classified as high-risk for ocelots also identified as high-risk for bobcats. Additionally, we used long-term felid population monitoring to evaluate effectiveness of recently constructed mitigation measures. We used six years of remote camera data from monitoring a road improvement project to evaluate mitigation efforts in the context of local felid abundance. Variation in seasonal abundance across the monitoring period was tied to precipitation, highlighting the importance of accounting for factors affecting wildlife abundance at mitigation sites when evaluating structure efficacy. Crossing structure use was positively correlated with abundance estimates, though frequency and consistency of use varied by individual. Probability of use was related to roadside fencing length, canopy cover, time since construction, and water in the crossing structure. For ocelots and other fragmentation-intolerant carnivore species, structures and fencing in focused areas where roads and intact habitat intersect may be feasible options to enable safe road crossing and promote connectivity.