In the context of a global crisis, where over 80% of children fail to meet minimum standards of physical activity, children with developmental disabilities may face an even greater health disparity. Despite being disproportionately less physically active than the broader population, those with developmental disabilities are often excluded from mainstream interventions. This study evaluated a novel animal-assisted intervention (AAI) that incorporated the family pet dog, with the goal of increasing physical activity in children with developmental disabilities. The intervention involved teaching child participants to engage in “Do as I Do” (DAID) reciprocal-imitation training with their dog over ten 60-min sessions. After enrollment, 45 child-dog pairs of children and their family dogs were randomly assigned to either the experimental intervention (DAID) group, the active control group (dog walking control [DWC]), or the passive control group (waitlist control [WC]). Physical activity was measured using a 7-day ActiGraph accelerometer monitoring process, and data were analyzed for children who wore the accelerometer for a minimum of 4 days at both assessment time points (N = 14). Participation in the intervention (DAID) group was found to significantly increase moderate-to-vigorous physical activity (MVPA) by 1.2 min/h (p = 0.0087; Cohen’s d = 1.7) and significantly reduce sedentary behavior by 4.03 min/h (p = 0.027; d = 1.3), outperforming the waitlist control group. Our findings demonstrate the potential of this novel dog-assisted intervention to improve physical activity in this population, thereby contributing to a reduction in health disparities.