Poor peer relations have been shown to be related to a variety of negative outcomes, including delinquent tendencies and poor academic performance. Researchers have found that coercive parent-child interactions are one precursor to this negative pathway. The development of children’s positive peer relationships is less clear, however, and it is equally important to understand children’s positive peer interactions. The purpose of this report is twofold. First, the nature of 7-year-old children’s positive peer relations is examined using a multi-method, multi-agent approach. A confirmatory factor analysis suggested four factors clustered by reporting agent and method: parents’ perceptions of their child’s friendships and social skills, teachers’ perceptions of the child’s positive attributes, observer impressions of the child-peer interaction in three play settings, and observer impressions of the child-peer interaction in a teaching context. Similar pathways were found for boys and girls. Second, once it was established that stable constructs could be developed to measure positive peer relations, earlier parent-child characteristics were examined to assess their utility in predicting children’s positive friendships. Parenting behavior was measured at 18 months and 5 years of age, and children’s attachment style was measured at 15 months of age. Parenting behaviors from the 18-month and 5-year assessments were combined to form a guided participation construct and a positive behavior construct using a confirmatory factor analysis. Results from a series of regression analyses indicate that observed parent positive behavior was a significant predictor of the child-peer interaction in the teaching context, and of the teacher ratings of the child. For girls only, attachment style was related to parent positive behavior.