AbstractGrowing human-wildlife conflicts and legal conservation obligations increased the need for precise information on Eurasian otter (Lutra lutra) population parameters for species status assessment and wildlife management measures. Scat surveys have become the method of choice to monitor species distribution range, abundance and habitat use. Although methodological concerns exist, scat abundance is often used as main indicator for otter population estimates and trends. To evaluate whether scat counts serve as good proxy for otter densities in linear stream habitats, we modelled the relationship between old/fresh scat abundance and otter numbers on two spatial scales, including also seasonal effects. Actual otter densities and marking behaviour were ascertained by genotyping of otter faeces collected at 218 marking sites along 132.6 km of four salmonid streams. Otter densities in our study areas ranged from 0.16 to 0.28 otters per stream kilometre. The age of scats and the level of spatial scale were determining factors when evaluating the suitability of scat abundance as index for otter densities. Covering large parts of an otter’s territory, fresh scat abundance increased with otter density, irrespectively of season. On the spatial scale of single marking sites, the relationship was not well supported, and no relationship could be found when using old scats in the models. Otter marking behaviour and differences in marking site use intensity, as well as seasonal differences in traceability and accumulation of old scats were discussed to bias the relationship between scat abundance and otter numbers.