A technique was developed for studying the reinforcement of one organism by another. Two pairs of monkeys served as subjects in adjoining but separate lever-pressing chambers. However, they were in visual, aural, and tactile contact with each other. After both pairs were trained to tolerate delays of reinforcement and one pair was trained under stimulus control to exchange reinforcements, monkey A of each pair pressed a lever to feed monkey B, and monkey B pressed to feed monkey A. The experiment sought to determine if this social interaction could be maintained. With a free responding procedure where the monkeys could work at any time in any order, the social relation proved unstable. After several oscillations in which one monkey did most of the responding and the other monkey did most of the eating, the reinforcement frequency for both pairs of animals decreased to very low levels. The final outcome would have been starvation had the experimenter not intervened.