Abstract Background A significant proportion of patients experiences poor response (i.e. no or little improvement) after total knee arthroplasty (TKA) because of osteoarthritis. It is difficult to quantify the proportion of patients who experiences poor response to TKA, as different definitions of, and perspectives (clinician’s and patient’s) on poor response are being used. The aim of this study was therefore to review the literature and summarize definitions of poor response to TKA. Methods A systematic search was performed to identify and review studies that included dichotomous definitions of poor outcome after primary TKA. The type, amount and combination of domains (e.g. functioning), outcome measures, type of thresholds (absolute/relative, change/cut-off), values and moments of follow-up used in definitions were summarized. Results A total of 47 different dichotomous definitions of poor response to TKA were extracted from 2163 initially identified studies. Thirty-six definitions incorporated one domain, seven definitions comprised two domains and four definitions comprised three domains. Eight different domains were used in identified definitions: pain, function, physical functioning, quality of life (QoL), patient satisfaction, anxiety, depression and patient global assessment. The absolute cut-off value was the most common type of threshold, with large variety in value and timing of follow-up. Conclusions Our inventory review shows that definitions of poor response to TKA are heterogeneous. Our findings stresses the need for an unambiguous definition of poor response to draw conclusions about the prevalence of poor-responders to TKA across hospitals and countries, and to identify patients at risk.