There is a growing body of evidence on the effectiveness of modified schema-based instruction (MSBI) for teaching word problem-solving alongside other important skills, such as self-determination and mathematical discourse and reasoning. Yet the majority of studies have not been conducted in naturalistic instructional arrangements. We report single-case design quantitative data from a mixed methods study that sought to evaluate the effectiveness and social validity of peer-delivered MSBI with high school students who were engaged in an established peer mentoring program. Results from the single-case multiple probe across participants design demonstrate a functional relation between peer-delivered MSBI and math problem-solving behaviors. All mentees with intellectual disability independently solved multiplicative comparison word problems following intervention and most maintained and generalized to a socially significant degree. Mentors delivered the intervention as designed and supported their mentees in expressing reasoning to justify their solutions. We provide recommendations for research and practice aimed at increasing expectations and opportunities for secondary students with intellectual disability regarding the use of peer-mediated interventions for academic skills.