Some token reinforcement programmes in schools have made tokens contingent upon the behaviour of the individual; others have made tokens contingent upon the behaviour of a group or groups. While both have been shown to be successful in controlling the target behaviours, there is no conclusive data avaiiable which would indicate which is the more effective contingency. This study was undertaken to make a comparison between these two reinforcement contingencies. The subjects were a class of seven pupils from a Christchurch Residential School for emotionally disturbed children. The dependent variable was disruptive behaviour. This was classified under seven categories for the pilot study and under five for the study proper. Data was collected during a daily half-hour Math period. Tokens were made contingent firstly upon the behaviour of the individual; and secondly upon the behaviour of the group. Preceding and following each of these contingencies there was a return to baseline conditions. It was found, that in this case, it was more effective to make the tokens contingent upon the behaviour of the group.