Poor sleep quality and low behavioral functioning are commonly reported in children with autism spectrum disorder. This study examined the impact of exercise on sleep on melatonin level and behavioral functioning in the population. Children with autism spectrum disorder (n = 55; age = 10.97 ± 1.90) were randomly allocated to a morning jogging intervention group or a control group. Participants' sleep was measured using actigraphy and sleep log assessments. Twenty-four-hour and first morning urinary 6-sulfatoxymelatonin were used to determine whether the exercise intervention could elicit changes in melatonin levels. Behavioral functioning of the participants was assessed by the repetitive subscale of the Gilliam Autism Rating Scale--3rd edition. All assessments were carried out in baseline, post-intervention, or regular treatment, and follow-up to elucidate the sustainability of the exercise effects. Positive changes were observed between baseline and post-intervention in actigraphy-assessed sleep efficiency and wake after sleep onset, as well as melatonin level and behavioral functioning within the intervention group (ps < 0.017). However, no significant changes were observed in all measurements between post-intervention and follow-up (ps > 0.05). The findings suggest that physical exercise is effective to improve sleep with an increase in melatonin level. It can also reduce repetitive behaviors in children with autism spectrum disorder.