Attachment representations are thought to provide a cognitive-affective template, guiding the way individuals interact with unfamiliar social partners. To examine the neural correlates of this process, we sampled event-related potentials (ERPs) during exclusion by unfamiliar peers to differentiate insecure-dismissing from securely attached youth, as indexed by the Child Attachment Interview. Thirteen secure and 10 dismissing 11-to 15-year-olds were ostensibly connected with two peers via the Internet to play a computerized ball-toss game. Actually, peers were computer generated, first distributing the ball evenly, but eventually excluding participants. Afterward children rated their distress. As in previous studies, distress was related to a negative left frontal slow wave (500-900ms) during rejection, a waveform implicated in negative appraisals and less approach motivation. Though attachment classifications were comparable in frontal ERPs and distress, an attachment-related dismissal dimension predicted a negative left frontal slow wave during rejection, suggesting that high dismissal potentially involves elevated anticipation of rejection. As expected, dismissal and self-reported distress were uncorrelated. Yet, a new approach to quantifying the dissociation between self-reports and rejection-related ERPs revealed that dismissal predicted underreporting of distress relative to ERPs. Our findings imply that evaluations and regulatory strategies linked to attachment generalize to distressing social contexts in early adolescence. © 2012 American Psychological Association.