This commentary focuses on the theoretical issues raised in this special issue. It is argued that both selection and influence in peer social networks reflect a process of peer contagion. The findings from the special issue studies are briefly reviewed. Some suggestions for future research are made, including the use of multilevel models that integrate interpersonal dynamics with peer network change as well as the use of natural and randomized experimentation to test hypotheses regarding the influence of contexts such as schools on the formation of peer networks. It is hoped that research on peer network can eventually be applied to the design of environments that promote a variety of health behaviors and peer support.