The road system to Buenaventura port (1946) and its interaction with wildlife was studied along its two legs 1. Unpaved road, low traffic volume (90kms; 76 vehicles/24hrs), low speed, and 2. Paved road, with heavy traffic at high volumes (126kms; 768 vehicles/24hrs), medium speed. We conducted a two phase research project that included geographical information system (GIS) corridor modeling tools supported by road data on drainage structures and their potential use as crossing structures, landscape forest patch size connectivity, and road kill data. Least-cost path corridors showed four major crossing sites, two on each road, for nine mammalian species present in three forest patches (>30km) in the area. Identified crossing hotspots correspond to associated forest patches along the roads. The 3,629 km surveyed evidenced a total of 19 species use the drainage structures for crossing, 38 species (164 individuals) were recorded as road kill and we assume 25 species of mammals present are also disturbed by the existing road. Both roads show differential impact on wildlife. Animals use partially open, leveled and well-lighted structures for crossing. Domestic animal species and surprisingly, human homeless or vagrants also used said structures. The lack of good wildlife crossing structures, human use of suitable wildlife underpasses as shelters, along with increasing human occupation are causing avoidance and causing fragmentation. The unpaved road has lower impact on wildlife but merits proper zoning for animal connectivity. Accelerated human disturbance coupled with poor road planning on the paved road system call for mitigation strategies such as domestic animal control, drainage structure maintenance and human use restrictions. Identified hotspots merit immediate traffic speed controls and animal underpasses in eventual road improvements and upgrades. Both roads have long established their effects on wildlife and new strategies will provide beneficial for resilient species still present close to the road.