OBJECTIVE: To systematically review the literature on the relationship between markers of inflammation and pain in patients with knee osteoarthritis (OA). METHODS: We searched MEDLINE, Web of Science and EMBASE databases from inception until June 2021. Eligible articles had to report on the association between inflammation (as measured by effusion, synovitis, baker's cysts, cytokines and C-reactive protein) and pain in patients with radiographic knee OA. Two reviewers independently performed a screening on title and abstracts, data extraction and risk of bias assessment using the Newcastle-Ottawa Scale (NOS). A best evidence synthesis was conducted for each inflammatory sign included in this review. RESULTS: 37 studies were included. Articles reported on the following measures: effusion or synovitis assessed via ultrasound (n = 9) or magnetic resonance imaging (MRI) (n = 17); baker's cyst (n = 3); cytokine concentrations (n = 11); and C-reactive protein levels (n = 4). The strength of the association between inflammation and pain does not exceed the moderate level (i.e., correlation coefficient values ranging from 0.19 to 0.61). Moderate levels of evidence were found for the association between synovitis (measured with ultrasound or contrast enhanced MRI) and pain. The levels of evidence between effusion (assessed via ultrasound), effusion/synovitis (assessed via non-contrast enhanced MRI), Baker's cyst, cytokines, C-reactive protein and pain were conflicting. CONCLUSIONS: Different inflammatory markers are associated with pain but the correlation ranges from weak to moderate, and the quality of evidence from conflicting to moderate. Further research is needed to strengthen the level of evidence and to establish mechanisms.