Although it is critical for students with autism to be included in large-scale assessment and accountability systems, it is not clear how to best measure their underlying academic skills and knowledge. Additional empirically-supported guidance is necessary to assist school teams that need to make decisions about how to best include students with autism in such systems. Test accommodations may be needed to ensure appropriate measurement on large-scale tests, but it is important to empirically explore whether they are meeting this intended purpose. In the current study, existing statewide test data were analyzed to determine the extent to which measurement comparability was apparent for students with autism receiving and not receiving accommodations on a math test used for accountability purposes. Limited differential item functioning (DIF) was identified for all comparisons, suggesting that existing school-based decisions to include students with autism in regular accountability testing do not correspond to serious measurement concerns. Implications for research and practice are provided.