Abstract The Eurasian otter Lutra lutra is a top predator that feeds mostly on fish. Yet, studies show the high plasticity of otter foraging behaviour and the varying importance of alternative food types in the species' diet. The aims of this study were, firstly, to demonstrate by meta‐analysis how the diet composition of otters varies depending on the habitat occupied, in terms of the types of land use, waterbody, and bank vegetation. Secondly, we tested whether the choice of method of presenting prey percentages in the diet of otters influenced the conclusions of the meta‐analysis. The percentage of fish in the diet was greater for otters occupying standing water than for those occupying flowing water; was higher for otters in waterbodies overgrown by reed vegetation; and varied depending on the type of surrounding land use. However, the results of the meta‐analysis varied depending on which of the two most common methods of diet analysis were used: percentage biomass or relative frequency of occurrence (RFO). We showed that the plasticity of otter feeding behaviour might be explained by the various habitats occupied by otters. Habitat differences reflected in otters' diets might have importance for the maintenance and conservation of local populations of this species. The high diversity of methods used by researchers to report the results of dietary studies precludes full comparisons and synthesis. This demonstrates the importance of unification of reporting. Moreover, the conclusions reached by the authors of any particular study might be highly dependent on the method they applied. Thus, care should be taken when comparing ecological studies based on different methods.