BACKGROUND: The Outcome Measures in Rheumatology (OMERACT) initiative developed a core outcome set (COS) of domains to assess effectiveness of interventions for knee osteoarthritis. These domains (pain, physical function, patient global assessment, imaging at 1 year) should be assessed in every trial to make research evidence meaningful and comparable. We systematically evaluated and critically appraised the use of measurement instruments and outcome domains in prospective studies evaluating patients with knee osteoarthritis undergoing total knee arthroplasty (TKA) and assessed their accordance with the OMERACT COS. METHODS: Literature search was performed until August 26, 2014, in Medline and Embase. Clinical trials and prospective observational studies with ≥50 participants and a follow-up of ≥1 year were included. We collected general study characteristics, comprehensive information on measurement instruments, and corresponding domains used. RESULTS: This systematic review identified low accordance of used outcome domains with the OMERACT COS of domains published in 1997. Only 4 of 100 included studies included all recommended core domains. Pain (85% of studies) and physical function (86%) were assessed frequently, whereas patient global assessment (21%) and joint imaging (≥1 year; 27%) were rarely assessed. There was substantial heterogeneity in the use of measurement instruments (n = 111) investigating TKA. CONCLUSION: More efforts are required to implement the existing COS. In addition, a more consistent use of adequate measurement instruments is important to make research evidence on TKA more relevant, better comparable, and thus more useful for guideline developers and clinical decision makers.