Abstract Background The treatment of knee osteoarthritis, which is a major cause of disability among the elderly, is typically selected from multidisciplinary options, including complementary and alternative medicine. Moxibustion has been used in the treatment of knee osteoarthritis in Korea to reduce pain and improve physical activity. However, there is no sufficient evidence of its effectiveness, and it cannot therefore be widely recommended for treating knee osteoarthritis. We designed a randomised controlled clinical trial to evaluate the effectiveness, safety, cost-effectiveness, and qualitative characteristics of moxibustion treatment of knee osteoarthritis compared to usual care. Methods/designs This is a protocol for a multicentre, pragmatic, randomised, assessor-blinded, controlled, parallel-group study. A total of 212 participants will be assigned to the moxibustion group (n = 106) and the usual care group (n = 106) at 4 clinical research centres. The participants assigned to the moxibustion group will receive moxibustion treatment of the affected knee(s) at 6 standard acupuncture points (ST36, ST35, ST34, SP9, Ex-LE04, and SP10) 3 times per week for 4 weeks (a total of 12 sessions). Participants in the usual care group will not receive moxibustion treatment during the study period. Follow-up will be performed on the 5th and 13th weeks after random allocation. Both groups will be allowed to use any type of treatment, including surgery, conventional medication, physical treatment, acupuncture, herbal medicine, over-the-counter drugs, and other active treatments. Educational material that explains knee osteoarthritis, the current management options, and self-exercise will be provided to each group. The global scale of the Korean Western Ontario and McMaster Osteoarthritis Index (K-WOMAC) will be the primary outcome measurement used in this study. Other subscales (pain, stiffness, and function) of the K-WOMAC, the Short-Form 36v2 Health Survey, the Beck Depression Inventory, the Physical Function test, Patient Global Assessment, and the Pain Numerical Rating Scale will be used as outcome variables to evaluate the effectiveness of moxibustion. Safety will be assessed at every visit. In addition, an economic evaluation and a qualitative study will be conducted as a mixed-methods approach. Discussion This trial may contribute to developing evidence for the effectiveness and safety of moxibustion for treating knee osteoarthritis. Trial registration Trial registration number:KCT0000130