Computation is foundational to learning many mathematics concepts, as well as a functional skill in everyday life. Yet students with autism spectrum disorder (ASD) often have challenges in learning computation skills. The current study aimed to provide quantitative and descriptive analyses of single-case experimental studies on computation interventions for students with ASD. Our systematic search and screening procedures yielded a total of 16 eligible studies for the final analyses. An analysis of publication bias indicated an absence of publication bias in the included studies. Tau-U was used to calculate the overall intervention effect, resulting in an omnibus effect of 0.90 (95% CI [0.81, 1.00]), indicating a medium effect. Moderator analyses showed that variables such as study design, participant characteristics, intervention characteristics, and target skills did not moderate intervention effects. However, some evidence-based practices were identified to significantly moderate maintenance effects. Implications for researchers and practitioners are discussed.