The purpose of this study was to determine whether the Penn Interactive Peer Play Scale (PIPPS), a teacher-rating instrument of interactive play behaviors for early childhood, was valid for urban, low-income children in kindergarten. The PIPPS demonstrated construct validity, and yielded three dimensions of interactive peer play: Play Interaction, Play Disruption, and Play Disconnection. These constructs were congruent with the dimensions found for preschool children. Concurrent validity was demonstrated with a standardized instrument assessing global social skills and academic competence. Children who displayed highly interactive peer play were given high ratings by teachers for social skills and were ranked higher in the class for academic competence. Those children who were disruptive or disconnected in play were viewed by teachers as having more problem behaviors and had lower academic achievement as compared to their peers. The PIPPS also was found to have predictive validity to first grade academic performance. Children who were reported by their teachers to have effective peer interactions during play had higher teacher ratings of academic success than children who were considered disruptive or disconnected in play. Implications for policy and practice in early childhood are discussed. (Contains 5 tables and 2 footnotes.)