Manipulatives are widely considered an effective practice and have been recommended as an evidence-based practice for students identified with a learning disability when used within the concrete--representational--abstract instructional framework. The aim of the current study was to evaluate single-case experimental designs that implemented a mathematics intervention using manipulatives on the mathematical outcomes of students at risk or identified with a disability. A total of 53 studies were included in the review. The Tau-"U" effect size (ES) across studies ranged from 0.34 to 1.00, with an omnibus ES of 0.91 (CI[subscript 95] = [0.87, 0.95]). The between-case standardized mean difference for individual studies ranged from 0.03 to 18.58. Moderator analyses revealed that out of nine variables analyzed (i.e., study quality, design, age, interventionist, manipulative type, perceptual richness, math concept, dependent variable, and disability category), only disability category served as a moderator. Implications for research and practice are discussed.