The current study examines the effects of schema instruction on the problem-solving performance of four second-grade students with emotional and behavioral disorders. The existence of a functional relationship between the schema instruction intervention and problem-solving accuracy in mathematics is examined through a single case experiment using a multiple baseline across participants design. Visual analysis and a non-parametric effect size (ES) demonstrate improvement in problem-solving accuracy for grade-level word problems involving addition and subtraction of two-digit integers without regrouping. Tau-U ESs for four participants ranged from 63% to 98%. The students and special education teacher reported the intervention package was socially valid.