Context. Mediterranean areas offer a mosaic of favourable microhabitats to reptiles (e.g. open zones, thorny bushes) and are considered as biodiversity hotspots for these organisms. However, in these dry and hot environments, reptiles remain sheltered most of the time. They generally escape observation, posing difficulties to perform inventories. Trap sampling or rock-turning surveys commonly used to detect reptiles entail important logistical constraints, may perturb fragile microhabitats, and are not appropriate for chelonians. Alternative simple and cost-effective methods are desired.

Aims. We tested the efficiency of camera trapping in a dry Mediterranean landscape, notably to detect threatened Hermann’s tortoises. We tested whether small artificial freshwater ponds could attract animals in the field of view of the cameras to increase detectability. We also tested whether sand tracks survey around ponds could improve the method.

Methods. We used a small number of cameras with ponds (5 in 2011, 7 in 2012), thereby maintaining low logistical costs. We randomly filled three ponds and emptied three ponds every 7 days. We set the time-lapse function of each camera with an interval of 5 min and inspected the sand tracks every 2 or 3 days. We used information from 39 radio-tracked tortoises to better estimate the detectability performances of the camera–pond system.

Key results. This technique was effective to detect tortoises (n = 348 observations) and five other reptiles (among the 11 species present in the study area). Large numbers of birds and mammals were observed (n = 4232, n = 43 species at least), thereby increasing the biodiversity list of the surveyed area. We detected 28% of the radio-tracked tortoises present in the monitored area. Filled ponds were more attractive and sand track survey completed camera monitoring.

Conclusions. Camera trapping associated with small ponds represent a useful tool to perform rapid inventories of the fauna in Mediterranean habitats, especially to detect the emblematic Hermann’s tortoise and other cryptic reptiles (e.g. snakes).

Implications. The low cost–efficiency ratio of this method allows performing multiple counting surveys, and thus may help collect robust data necessary to justify the protection of key habitats that are coveted by property developers.