[Objectives] Infrared camera is a classic wildlife observation method with many advantages, including continuity, non-invasiveness, etc. Therefore, it has been widely used in monitoring important wildlife species like Giant Panda (Ailuropoda melanoleuca). Giant Panda acts as a flagship species and an umbrella species in Qinling. According to the fourth national survey on Giant Pandas, its population has picked up during 2003-2011 in the Qinling Mountains with the contribution of local ecological recovery. Qinling is not only a major habitat of Giant Panda nowadays, but also holds a Giant Panda population evolutionarily closer to the common ancestor of Giant Panda. Monitoring the Giant Panda population and biodiversity in its habitats is therefore essential for a better understanding towards its distribution, living patterns and protection. This research aims to investigate the wildlife diversity and activities during the Giant Panda population recovery period with prohibition of human activity, and use trapping images and diversity indexes to discuss similarities and differences between the wildlife distribution in the middle Qinling Mountains. [Methods] From August 2018 to October 2020, 80 infrared cameras were deployed among 4 valleys (Danangou Valley, Ponianzigou Valley, Dagupingyanwugou Valley and Longtanziyanwugou Valley) located in 3 nature reserves (Huangbaiyuan Nature Reserve, Guanyinshan Nature Reserve and Foping Nature Reserve) to monitor wildlife diversity and activities in the population recovery region of Giant Panda in the middle Qinling Mountains (Fig. 1 and 2, Table 1). All photos and videos were examined, and individual photos and individual richness of each species were counted. Species diversity on overall data (also for mammals and birds separately) and among different valleys was investigated using relative abundance index, rank abundance curve, Shannon-Wiener index and evenness index. Species similarity indices between valleys were calculated. All analyzing methods above were also used separately on data collected in valleys and on ridges. The rank abundance curves were tested by R-squared coefficients. [Results] A total of 34 514 bird and mammal images have been collected during the monitoring period, with a total effective working days of 54 240 d. 12 364 individual photos are identified and 10 612 of them can be confirmed into 68 species with 21 mammal species and 47 birds (Appendix 1). Based on this, a wildlife image database is established, with which the species diversity in the monitoring area is analyzed. Reeves’s Muntjac (Muntiacus reevesi), Wild Boar (Sus scrofa) and Tufted Deer (Elaphodus cephalophus) are the three most abundant mammals, and Golden Pheasant (Chrysolophus pictus), Greater Necklaced Laughingthrush (Garrulax pectoralis) and White-throated Laughingthrush (Garrulax albogularis) are the three most abundant birds. The rank abundance curve of birds is flatter with a longer tail compared to mammals’ (Fig. 3). The Shannon-Wiener index of birds is higher than that of mammals, whilst the evenness index is lower (Table 2). Species are not evenly distributed among valleys. Danangou Valley in Huangbaiyuan Nature Reserve obtains the most individual photos (28.04%), while Longtanziyanwugou Valley in Foping Nature Reserve has the smallest proportion of individual photos (21.75%) (Fig. 4). Shannon-Wiener index and evenness index indicate an overall highest species diversity and evenness in Longtanziyanwugou Valley and the lowest diversity and evenness in Ponianzigou Valley in Guanyinshan Nature reserve, and a comparably high bird diversity and evenness in Dagupingyanwugou Valley in Foping Nature Reserve (Table 3). Species similarity indices of mammals among 4 valleys are significantly higher than those of birds. Ponianzigou Valley and Longtanziyanwugou Valley share the highest species similarity, while Ponianzigou Valley and Dagupingyanwugou Valley have the lowest species similarity (Table 4). Cameras placed on ridges identify more species and individual photos than those in valleys (Fig. 5). Higher Shannon-Wiener index and evenness index of all species and mammals are also found on the ridges, but the trends of birds are reversed, which might because of their preference of dense forests (Table 5). Species similarity indices between each ridge and valley indicate that species are more similar among different ridges than among different valleys, suggesting potential geographic isolation between valleys (Fig. 6). [Conclusion] In conclusion, this study enriches the basic biological database in middle Qinling Mountains, as well as provides data support for further scientific research and ecological protection policy. The biodiversity and species evenness are generally high in research areas. Birds are more diverse (with more species) than mammals but with less evenness and individual photos. Ponianzigou Valley has the lowest biodiversity and evenness among the 4 valleys in accordance with its shortest ecological recovery timespan, which calls for further attention. While Longtanziyanwugou Valley has the highest species diversity and evenness, although its individual photos are the least, which needs further investigation. Dagupingyanwugou Valley has a low mammal activity level, and more targeted protection measures are required. A total of 37 individual photos of Giant Panda were shot, with over 80% in Danangou Valley and none in Longtanziyanwugou Valley, reflecting substantial ecological recovery in Danangou Valley.