The present investigation examined observational learning during equivalence-based instruction using the PEAK Equivalence curriculum, across five children with disabilities. A multiple baseline across skills was used to evaluate observational learning across the participants, including directly observed relations, as well as derived symmetrical and transitive relations. Each skill contained 5 stimulus classes, where a subset of relations for a single class was trained for each participant. The remaining 4 class subsets were observed by each of the participants. For the derived relations, participants never observed peers contact reinforcement for correct responding nor did they themselves contact reinforcement for correct responding. After multiple sessions, results indicated increases in both observed directly trained relations, as well as emergent derived relations. Over the course of the sessions, the participants average correct answers increased from an average of 12% during baseline to100% at the end of the training phase and during a follow-up test probe. The results have implications for equivalence-based ABA instruction in schools and other settings where group-based instruction is common in application with individuals with disabilities.