摘要:This methods report illustrates the relevance of Mendeley readership as a tool for research assessment. Readership indicators offer new possibilities to inform the evaluation of publications and outputs either poorly covered in citation indexes (e.g. non-English language outputs, Global South publications, Social sciences and humanities), or typically excluded from citation analysis (e.g. letters, editorial material, etc.). Mendeley readership can also inform the earlier impact of scientific outputs, as well as the impact among wider non-academic audiences. All these features are discussed in this report and the relevance of readership indicators to extend the concept of research impact beyond specific acts (e.g. citations) is highlighted. Best practical recommendations on how Mendeley readership can be used for assessment purposes are discussed. Policy highlights: This paper illustrates practical possibilities of readership indicators for research evaluation. Readership indicators inform impact of publications poorly covered in bibliometrics databases or excluded from citation analysis. Readership indicators inform early impact and non-academic impact of publications. Readership indicators can be used to inform, support, and complement (citation-based impact) decisions on research evaluation exercises.