<p>The monitoring of wildlife populations is a central aspect for conservation and wildlife management. As funding is often limited, monitoring is required to produce data of sufficient accuracy while being implemented with the least cost as possible. In this study, we investigated the use of hand-held thermal cameras for their application in surveys of the hazel dormouse <i>(Muscardinus avellanarius)</i> and the edible dormouse <i>(Glis glis).</i> We compared the detection probability using this novel method with an approach using tracking tunnels which were set in bushes and on the ground. Between 2021 and 2022, transects of 200&#xa0;m were established at 18 sites in the Jura mountains in Switzerland. A transect consisted of 8 to 10 tracking tunnels and were each set once between July and September. Eight of these 18 sites were then visited with a hand-held thermal camera during the time when the tracking tunnels were set. For hazel dormice the detection estimate for thermal camera for a single visit was 0.82 while the detection probability for a tracking tunnel in the bushes was 0.05 and for one on the ground 0.005. For the edible dormouse, one single visit using the thermal camera had a detection probability of 0.68, for a tracking tunnel in the bush 0.03 and on the ground 0.05. Our results show that thermal cameras can be used to detect dormice species and are highly effective compared to tracking tunnels. However, further investigations are required to implement thermal imagery in monitoring the presence of these animals. This includes effects such as season, habitat, time, duration and repetitions of transects as well as the validation of the observations.</p>