Social play between pairs of individually housed juvenile rats was studied by measuring pinning frequencies occurring during rough-and tumble play. Low doses of morphine increased play, and opioid blockade with naloxone reduced play. Dominance, as measured by which pinned which most, was markedly reduced by naloxone and slightly increased by morphine. These effects were most apparent when animals were treated with these drugs continuously from the beginning of play testing. Although the respective drugs tended to reverse established dominance patterns (i.e., naloxone made dominant animals more submissive, and morphine made submissive animals more dominant), it was apparent that previously established social learning limited the effectiveness of these manipulations. In general, the results are consistent with the proposition that brain opioids are important in controlling the vigor of social relations.