OBJECTIVE To determine the effects of magnetic knee wraps (MKWs) on joint position sense (JPS) in individuals with mild-to-moderate knee osteoarthritis. METHODS Randomized, double-blind, placebo-controlled pilot trial. METHODS Rehabilitation clinic of a tertiary hospital. METHODS Fifty volunteers with mild-to-moderate knee osteoarthritis participated and 33 (66%) completed the trial; only one (2%) withdrew due to study-related adverse effects. METHODS Knee wraps with a static magnetic field (35  mT) were compared with control knee wraps for 12 weeks. METHODS The main outcome measure was JPS at first, fourth, and twelfth week while applying of the knee wrap, quantified as the ability to replicate target knee joint angles. The secondary outcome was the Comprehensive Osteoarthritis Test (COAT) total score. RESULTS In the intention-to-treat analysis with an alpha-level of 0.05, 12-week application of the MKWs did not significantly reduce the mean absolute error (MAE) of JPS (n = 18; P = 0.627); contrarily, the MAE was found to decrease significantly after wearing the sham MKWs (n = 20; P = 0.01). The COAT total score gradually improved in both groups (P < 0.05). When comparing the improvement of MAE and COAT total score at each follow-up between two groups, no significant differences were found; the maximal difference in the improvement of MAE between two groups was noted at week 12 (mean difference: 1.2°; 95% confidence interval -0.4 to 2.9; P = 0.144). In the per-protocol analysis, the study group showed better improvement in the COAT total score at week 12 (P = 0.033). CONCLUSIONS The 35-mT MKWs could not significantly influence knee JPS in individuals with mild-to-moderate knee osteoarthritis.