More than 800,000 students with autism spectrum disorder (ASD) are served in U.S. public schools, a number that has increased annually for almost 20 years since Individuals with Disabilities Education Act (IDEA) included ASD as an eligibility category. These students often experience widespread and persistent reading comprehension challenges. A systematic review and meta-analysis, anchored in a construction-integration model of reading comprehension, was undertaken to examine single-subject research design (SSRD) and group design studies aimed at remediating reading comprehension challenges in students with ASD. Reading comprehension instruction was operationalized by surface, textbase, and situation model levels. Random effect modeling with robust standard error estimations suggested small yet statistically significant intervention effects. Situation model-level interventions demonstrated the highest intervention effects in group design studies. Across designs, retelling and curriculum-based question-answering measures demonstrated larger effect sizes than cloze and norm-referenced measures. Methodological quality analysis suggested that few individual SSRD and group studies had high design quality with large intervention effects. We concluded that reading comprehension instruction focused on the situation model level should be prioritized for students with ASD, and future studies should reduce bias in designing intervention studies.