Objective To research the therapeutic effect and variation causes of TCM clinical pathway for knee osteoarthritis（KOA） and propose measures for improvement. Methods Sixty cases of KOA were treated with TCM clinical pathway. The scores of visual analogue scale（ VAS）

Lequesne index and Western Ontario and McMaster Universities Arthritis Index（WOMAC） were compared before and after treatment. The changes were closely observed according to the contents of TCM clinical pathway table. The reasons for variation were described in detail and the improvement measures were proposed. Results The total effective rate of TCM clinical pathway for KOA was 98. 3%. The scores of VAS

Lequesne index and WOMAC were significantly decreased after treatment（P < 0. 05 or P < 0. 01）. The positive variation was 11. 21% and the negative variation was 88. 79%. The sources of variation showed that medical personnel accounted for 28. 97%

patients accounted for 32. 71%

hospital systems accounted for 31. 78% and disease outcome accounted for 6. 54%. The controllable variation was 50. 47% and the medical personnel and hospital systems accounted for 28. 97% and 21. 50% respectively. The uncontrollable variation was 49. 53% and the patients

hospital systems and disease outcome accounted for 32. 71%

10. 28% and 6. 54% respectively. Conclusion The TCM clinical pathway for KOA is effective and it can obviously relieve syndrome. The reasons for variation are medical personnel

patients

hospital systems and disease outcome. The improvement measures are strengthening cooperation of hospital departments

training medical personnel

publicizing TCM clinical pathways and advocating rational medical therapy.