This study examined the validity of teacher assistant report of children's classroom peer play behaviors using the Penn Interactive Peer Play Scale (PIPPS). Participants for the examination of construct validity were 344 urban Head Start children. Exploratory factor analyses resulted in a reliable three-factor solution, corroborating the dimensions of interactive peer play previously found with teacher and parent reports. Factor-matching analyses yielded high congruence between the teacher-assistant structure and the credentialed-teacher and parent structures. Additionally, concurrent validity was demonstrated with a subsample of 270 children by significant relationships found between the PIPPS dimensions and teacher observations of children's classroom learning competencies and an independent assessment of children's early academic success. The results affirm that teacher assistants in Head Start are a reliable and valid source of information on children's interactive, peer play competencies. (Contains 2 tables.)