Research has yet to discount all sources of confounding in the relationship between an individual’s delinquent behavior and that of his or her peers. One approach is to control for an active gene–environment correlation (rGE). Active rGE occurs when one selects into an environment based on genetic propensities. The current study utilizes twin data from the National Longitudinal Study of Adolescent Health to examine the impact of a direct measure of peer delinquency on self-reported delinquency while controlling for active rGE. The final analytic sample ranged between 456 and 524 dizygotic and 286 and 350 monozygotic twins, depending on the measures being analyzed. Using an augmented version of the DeFries–Fulker model, results revealed the peer effect was no longer statistically significant once genetic confounding (active rGE) was controlled. These findings support selection arguments and run counter to learning theory explanations.