We conducted a contingency analysis to evaluate if a sequential relation between social initiations and positive social responses increased for both therapists and children with autism following a conditioned reinforcement procedure. Participants included child-therapist dyads, which were previously identified as having low rapport. These dyads were observed prior to and following an intervention designed to establish therapists' social behavior as a reinforcer. Sessions consisted of unstructured play between the therapist and child. Results from a Yule's Q analysis show that both the child and adult positive responding to the others' social initiations increased following the intervention. Findings highlight the reciprocal effects of therapist-child interactions as well as the effectiveness of establishing social attention as a reinforcer via an operant discrimination training procedure.