BACKGROUND: Tai Chi (TC) controls pain through mind-body exercise and appears to alter inflammatory mediators. TC actions on lipid biomarkers associated with inflammation and brain neural networks in women with knee osteoarthritic pain were investigated. METHODS: A single-center, pre- and post-TC group (baseline and 8 wk) exercise pilot study in postmenopausal women with knee osteoarthritic pain was performed. 12 eligible women participated in TC group exercise. The primary outcome was liquid chromatography tandem mass spectrometry determination of circulating endocannabinoids (eCB) and oxylipins (OxL). Secondary outcomes were correlations between eCB and OxL levels and clinical pain/limitation assessments, and brain resting-state function magnetic resonance imaging (rs-fMRI). RESULTS: Differences in circulating quantitative levels (nM) of pro-inflammatory OxL after TC were found in women. TC exercise resulted in lower OxL PGE1 and PGE2 and higher 12-HETE, LTB4, and 12-HEPE compared to baseline. Pain assessment and eCB and OxL levels suggest crucial relationships between TC exercise, inflammatory markers, and pain. Higher plasma levels of eCB AEA, and 1, 2-AG were found in subjects with increased pain. Several eCB and OxL levels were positively correlated with left and right brain amygdala-medial prefrontal cortex functional connectivity. CONCLUSION: TC exercise lowers pro-inflammatory OxL in women with knee osteoarthritic pain. Correlations between subject pain, functional limitations, and brain connectivity with levels of OxL and eCB showed significance. Findings indicate potential mechanisms for OxL and eCB and their biosynthetic endogenous PUFA precursors that alter brain connectivity, neuroinflammation, and pain. CLINICAL TRIAL REGISTRATION: ClinicalTrials.gov, identifier: NCT04046003.