The purpose of this review was to systematically search and critically analyse the literature pertaining to behavioural outcomes of exercise interventions for individuals with autism spectrum disorder aged <16 years. This systematic review employed a comprehensive peer-reviewed search strategy, two-stage screening process and rigorous critical appraisal, which resulted in the inclusion of 13 studies. Results demonstrated that exercise interventions consisting individually of jogging, horseback riding, martial arts, swimming or yoga/dance can result in improvements to numerous behavioural outcomes including stereotypic behaviours, social-emotional functioning, cognition and attention. Horseback riding and martial arts interventions may produce the greatest results with moderate to large effect sizes, respectively. Future research with well-controlled designs, standardized assessments, larger sample sizes and longitudinal follow-ups is necessary, in addition to a greater focus on early childhood (aged 0-5 years) and adolescence (aged 12-16 years), to better understand the extent of the behavioural benefits that exercise may provide these populations.