OBJECTIVES: Systematic reviewers utilise filters to focus searches of electronic databases to identify specific study designs. Established, tested search filters are available from groups which regularly conduct reviews such as the Cochrane Group, SIGN and CADATH. Although studies in the real-world are not new, novel phrases such as real-world evidence (RWE) are increasingly used to identify observational studies. Currently available search filters for observational and non-randomised studies do not adequately capture newer terms used. We have thus developed a method to identify frequently used MeSH terms in RWE studies and a search filter to include these terms. METHODS: A PubMed (MEDLINE®) search for RWE stated in the title/abstract was conducted. Articles with “real-world” and either “data”, “evidence” or “research” in the title or abstract were selected; Case Reports, Comment, Editorial, Letter, News were removed. MeSH terms associated with articles were analysed and frequency counted; those relating to study design or outcome reporting were chosen for inclusion in the search filter. RESULTS: The MEDLINE® search identified 179 studies reporting RW and either data, evidence or research in the title. Of the 179, 151 were publication types of interest. The most frequently used MeSH terms related to RWE identified were ‘Treatment Outcome’ (n=30), ‘Evidence-based Medicine’ (n=17), ‘Retrospective Studies’ (n=15), ‘Databases, Factual’ (n=14), and ‘Time Factors’ (both n=14). A search strategy was developed combining MeSH and free-text terms to identify RWE. CONCLUSIONS: For every systematic review it is important to validate searches to ensure they are retrieving relevant studies; as new terminologies such as RWE are introduced to describe study design, reviewers need to adapt search filters. The method proposed allows searches to be adapted as terminologies are introduced and become more established.