Objective To evaluate the test-retest reliability and validity of Chinese version of the Ottawa self-injury inventory. Methods The test-retest reliability and internal consistency reliability of Chinese version of the Ottawa self-injury inventory were evaluated by a questionnaire survey of 163 students and 20 clinical patients in Shanghai. Results The test-retest reliability coefficient of all items was above 0.400, except for the item of “do anything to keep hands busy” for the best way to resist the non-suicidal self-injury (NSSI). The overall test-retest reliability was satisfactory. The internal consistency Cronbach's α coefficient of function subscales was 0.952 and the internal consistency Cronbach's α coefficients of its 4 factors ranged from 0.637 to 0.896. Factor analysis revealed 6 factors, which accounted for 75.0% of the variance. Conclusion The test-retest reliability and construct validity of Chinese version of Ottawa self-injury inventory are satisfactory and construct can be used for investigating the behavioral characteristics and causes of NSSI.