From September 2017 to February 2018, 650 disabled children and adolescents aged 6-to 17-year-old in Beijing were matched 1∶1 to those in normal physical condition with same age, gender and place of residence. All children and adolescents were investigated for self-injury status in the last year to understand the difference of self-injury incidence between groups. Multivariate unconditional logistic regression model was applied for exploring relevant factors of self-injury of children and adolescents. This study found that the disability, insufficient sleep, difficulty falling asleep, and sleeping late were associated with self-injury of children and adolescents.