Background. Children with Autism Spectrum Disorder (ASD) have few opportunities to engage in physical activity, which affects their already limited social and motor skills. Different Service-Learning (SL) proposals designed to meet this need have started to emerge, and research is required to better understand the services provided. Objective. Two objectives were established in this study: (1) to understand the impact of a physical education SL programme from the viewpoints of Physical Education Teacher Education (PETE) students and the families of the children with ASD involved; and (2) to analyse the programme’s effects on the motor domain of children with ASD (physical activity levels and basic motor skills). Method. The study used a mixed-methods approach. Thus, for the first objective, a qualitative approach consisting of group interviews and reflective journals was used with a sample composed of 12 parents of the participating children with ASD and 52 PETE students. For the second objective, a quasi-experimental design was used with a sample of 25 children with ASD, 15 in the experimental group (4 girls, 11 boys; age: 10.13 ± 2.56 years) and 10 in the control group (1 girl, 9 boys; age: 10.13 ± 3.09 years). In this part, triaxial accelerometers (Actigraph, xGT3X-BT) and the Movement Assessment Battery for Children-2 (MABC-2) were used. Results. Five content categories emerged from the qualitative approach, three of which were shared by the families and the PETE students: ‘Motor-related benefits for children’, ‘Social impact regarding inclusivity’, and ‘Teacher training quality’. In addition, the ‘Personal growth’ category emerged from the PETE students, and ‘Desire for consolidation of the programme’ emerged from the families. Regarding the quantitative approach, the experimental group showed significant improvements in moderate Physical Activity performed on weekdays (p ≤ 0.001), as well as on the Manual Dexterity (p ≤ 0.05) and Balance (p ≤ 0.05) variables from the MABC-2 test. Meanwhile, the results in the control group remained stable or decreased. Conclusion. The SL programme implemented was beneficial for all the parties involved and successfully achieved the objectives established a priori. In any case, further research on programmes and services for children is needed to guide future proposals.