OBJECTIVE: The 6-minute walk test (6MWT), an Osteoarthritis Research Society International (OARSI)-recommended measure of physical function in knee osteoarthritis (OA), was originally developed to assess submaximal aerobic fitness in people with cardiovascular disease. The degree to which 6MWT performance reflects knee OA severity versus other patient factors remains unclear. Our objective was to assess the contributions of OA-related and non-OA-related patient characteristics to 6MWT performance in individuals with symptomatic knee OA. DESIGN: In this cross-sectional study, participants scheduled for total knee arthroplasty completed the 6MWT and standardized questionnaires. Participant characteristics were compared by tertiles of 6MWT distance. Multivariable linear regression modelling was used to assess associations between patient factors and 6MWT distance. RESULTS: Among 278 participants (mean age 67.1 years, 65.5% female, mean WOMAC pain 57.0, and mean 6MWT distance 323.1m), older age (adjusted beta coefficient -4.0 per year, 95% confidence interval [CI] -5.4, -2.5), female sex (adjusted beta [95% CI] -43.0 [-67.3, -18.7]), presence of obesity (adjusted beta [95% CI] -44.4 [-68.2, -20.5]), greater knee pain (adjusted beta [95% CI] per unit increase in WOMAC pain -1.12 [-2.1, -0.2]), worse knee-OA related function (adjusted beta [95% CI] per unit increase in KOOS-PS -1.1 [-2.0, -0.1]) were associated with shorter 6MWT distance. Greater arthritis coping efficacy (adjusted beta [95% CI] 5.2 [1.4, 9.0]) was associated with longer 6MWT distance. CONCLUSIONS: While 6MWT distance declines with greater knee OA symptom severity, other demographic, biomedical and psychosocial factors also significantly influence 6MWT performance. These should be considered when interpreting the 6MWT in this population.