The eastern area of the Yushan National Park abounds with wildlife with the artiodactyla and carnivora being the most prominent groups. Herbivory and trampling by artiodactyla would exert influence on vegetation, and their abundance and activity in turn are susceptible to human interference. Therefore, long-term monitoring of the population status and their effects on habitat are needed. The small to medium-sized carnivores were the least known wildlife in this area due to their elusive nature, nocturnal habits and low densities. Recently, carnivore tracks and signs became more easily seen along the trail. However, there was difficulty in species identification for most of the carnivore feces found in the field. This study aims to monitor the status of artiodactyla species and to discern the carnivore species inhabiting the area. The monitoring program was employed by sign counts and infrared-triggered cameras. Carnivore feces found during the survey were collected for genetic and diet analysis. Species were determined by the sequences of 12S rRNA gene of mitochondrial DNA extracted from feccal samples. Diet of carnivores were studied by fecal analysis. The results showed the amount of the signs of a given species paralells to the spread of signs along the trail. Tourists, heavy rain and mowing interfered the persistency of fecal pellets and activity of wildlife along the trail. Mustelid-type feces were more abundant than mongoose-type feces and the spatial and seasonal occurrence of the two types of feces varied. Daily capture rates by automatic cameras and rate of occurrence on camera stations were the highest for Muntiacus reevesii and Macaca cyclopis. Among carnivore species, camera capture rates were higher for Herpestes urva and Paguma larvata in Walami section and the camera capture rates of Martes flavigula were similar in the four sections. Sus scrofa was the least recorded species by either method, while less Paguma larvata and Melogales moschata were found by signs and tracks survey than by camera recording. Success rate in genetic species identification of the fecal samples was higher for Herpestes urva. However, DNA of the preys eaten by Martes flavigula were more readily extracted from feces. The molecular technique of genetic species identification employed in this study awaits further improvement. The technique can also be applied to ascertain the prey species for carnivorous mammals.