Surveillance of Echinococcus multilocularis at the edge of its range is hindered by fragmented distributional patterns and low prevalence in definitive hosts. Thus, tests with adequate levels of sensitivity are especially important for discriminating between infected and non-infected areas. In this study we reassessed the prevalence of E. multilocularis at the southern border of its distribution in Province of Bolzano (Alto Adige, northeastern Alps, Italy), to improve surveillance in wildlife and provide more accurate estimates of exposure risk. We compared the diagnostic test currently implemented for surveillance based on coproscopy and multiplex PCR (CMPCR) to a real-time quantitative PCR (qPCR) in 235 fox faeces collected in 2019 and 2020. The performances of the two tests were estimated using a scraping technique (SFCT) applied to the small intestines of a subsample (n = 123) of the same foxes as the reference standard. True prevalence was calculated and the sample size required by each faecal test for the detection of the parasite was then estimated. True prevalence of E. multilocularis in foxes (14.3%) was markedly higher than reported in the last decade, which was never more than 5% from 2012 to 2018 in the same area. In addition, qPCR showed a much higher sensitivity (83%) compared to CMPCR (21%) and agreement with the reference standard was far higher for qPCR (0.816) than CMPCR (0.298) meaning that for the latter protocol, a smaller sample size would be required to detect the disease. Alto Adige should be considered a highly endemic area. Routine surveillance on definitive hosts at the edges of the E. multilocularis distribution should be applied to smaller geographic areas, and rapid, sensitive diagnostic tools using directly host faeces, such as qPCR, should be adopted.