OBJECTIVES: There are a number of search filters designed to identify studies with particular study designs in electronic databases. This study compared three filters for identifying randomised controlled trials (RCTs). METHODS: Searches were conducted on 15^thJune 2015 in the Ovid MEDLINE and MEDLINE In-process databases using The Cochrane Highly Sensitive Search Strategy for Identifying Randomized Trials in MEDLINE, the SIGN Randomised Controlled Trials MEDLINE filter and the BMJ MEDLINE Randomised Controlled Trial Strategy. Differences were explored by reviewing samples of records uniquely identified by each filter. For comparison, a sample of articles returned by all three filters was also reviewed. To estimate the sensitivity of each filter, the detection of 39 publications of RCTs included in a randomly-selected Cochrane Collaboration systematic literature review (SLR) was tested. RESULTS: CONCLUSIONS: All filters failed to identify at least some RCTs, SLRs or meta-analyses of RCTs. Differences between filters, including the publications uniquely identified by each, should be considered when selecting filters for use in literature reviews.