Multiple non-invasive methods are used to monitor elusive carnivores but often rely on expensive scat-detection dogs or remote cameras. Walking surveys for field sign (tracks and scat) are an inexpensive alternative but may be less accurate. This study aims to calibrate occupancy estimates of ocelots (Leopardus pardalis) generated from field sign surveys against the occupancy estimates generated from remote camera surveys in the Greater Madidi-Tambopata Landscape of the Bolivian Amazon. Field sign surveys are expected to yield significantly lower overall occupancy estimates than remote camera surveys but generate a similar pattern of spatial use across the landscape, especially in areas with ideal substrate for tracks (sand and mud along streams). Thus, field sign surveys are predicted to be suitable for assessing ocelot presence or absence and determining target areas for further study as a supplement to remote cameras. In June, August, and September 2019, a 60-km2 grid of 60 camera stations was deployed for 40 days. Each station was paired with three 600 m transects for field sign walked by indigenous guides and experienced researchers. Ocelot occupancy at each station will be calculated from both the remote cameras and transects, and the results quantitatively compared. Testing the reliability of field sign surveys will allow carnivore monitoring programs to incorporate this low-cost method into their studies without sacrificing accuracy, as well as offer opportunities to integrate indigenous knowledge of animal tracking into conservation science.