Piaget (1932) hypothesized that children's interactions with peers during middle child hood are essential to their moral reasoning development. To test this hypothesis, 48 female focal subjects (M age = 8.6 years) were paired with either a female agemate or their mother. All focal subjects were pretested and posttested for moral reasoning abilities. In the intervention, the adult-child and peer dyads engaged in consensus-seeking discussions of two moral dilemmas. Focal subjects' moral reasoning at pretest and posttest and their use of reasoning (transacts) in the intervention discussions were measured. As predicted, focal subjects paired with peers showed significantly more sophisticated moral reasoning subsequent to their discussions than did focals paired with adults. In addition, focals paired with peers used more active transacts in their discussions than did focals paired with adults. Styles of dyadic discussion that featured active transacts by focal subjects were positively correlated with the focals' moral reasoning at posttest, whether the focal subject was paired with a peer or an adult. The more sophisticated posttest reasoning by focals paired with peers was attributed to the greater use of active discussion styles in peer dyads.