A meta-analysis of single-subject research was conducted examining the effectiveness of computer-assisted interventions (CAI) for teaching a wide range of skills to students with autism spectrum disorders (ASD) within the school-based context. Intervention effects were measured by computing improvement rate difference (IRD), which is a simple approach to visual analysis that correlates well with both parametric and non-parametric effect size measures. Overall, results suggest that CAI may be a promising approach for teaching skills to students with ASD. However, several concerns make this conclusion tenuous. Recommendations for future research are discussed.