Rodent predators are largely exposed to anticoagulant rodenticides (ARs) and, to mitigate their exposure, drivers of transfer should be better characterized. The measurement of ARs residues in faeces appears as a potential non-invasive indicator to assess vole predator exposure. However it is unknown whether ARs residues in faeces could be related to treatment patterns. In 2011, fox-like faeces were sampled in 2 contexts of ARs usage, "plant protection product (PPP)/biocide" or "biocide only". PPP treatments using bromadiolone were carried out to control water vole Arvicola terrestris populations. PPP treatments were quantified and located. In each usage, 160 faeces of vole predators were geo-referenced and then, stored at 20°C for further ARs titration (LC-ESI/MS/MS) and species identification (DNA-PCR). DNA was amplified for 37.2% of faeces. Among them, the most frequent species was the red fox (73.9%), then cat (21.8%) and finally dog (4.2%). ARs occurrence did not differ between fox, cat and PCR unidentified faeces (p=0.35). Every positive faeces contained only bromadiolone except one (biocide context) with chlorophacinone. ARs were detected more frequently where PPP treatments occurred (p<0.001). In PPP/biocide context, the ratio of positive faeces varied non-linearly with the area of PPP treatments within a 1km-radius around faeces (p<0.001; pseudo-r²=0.43). It increased exponentially up to 70% until a 0.4 km² area treated. Then this ratio increased almost linearly to reach 90% for 0.85 km2. Those results indicate that faeces collected in situ may be a relevant non-invasive target to monitor AR exposure comparatively.