Objective To test the reliability and validity of the famity inveritory of sleep habits(FISH) in the sleep assessment among children with autism spectrum disorder (ASD). Methods A random sample of 199 children with ASD and sleep disorders intervened in the Children's Language and Behavior Department of Maternal and Child Health Hospital of Qingyuan City were selected, their caregivers completed the FISH and the children's sleep habit questionnaire(CSHQ) at the same time, and completed the above two questionnaires again 2 weeks later to test the reliability and validity of the FISH. Results The results of confirmatory factor analysis showed that the five-factor model fits well (χ²=79.05, df=44, χ²/df=1.80, P<0.001, SRMR=0.06, GFI=0.94, IFI =0.89, RMSEA=0.06), the total scale and each subscale had high internal consistent reliability (0.71~0.77) and test-retest reliability (0.79~0.88). The daytime habits of FISH were negatively correlated with sleep delay and waking up at night of CSHQ respectively, the habits before sleep of FISH were negatively correlated with sleep conflict, sleep anxiety and daytime sleep of CSHQ respectively, the sleep routine of FISH was negatively correlated with sleep conflict, sleep delay and sleep anxiety of CSHQ respectively, the parents' behavior before sleep of FISH was negatively correlated with sleep conflict and sleep anxiety of CSHQ respectively, and the total score of FISH was negatively correlated with CSHQ, with statistically significant differences (P<0.05). Conclusions The FISH had good reliability and validity, and can be used as an assessment tool for children with ASD.