Fifteen overweight girls aged 5 to 11 yr were randomly assigned to one of two weight-reduction treatments: response-cost plus reinforcement, response-cost only, or a no-treatment control group. In the response-cost plus reinforcement group, parents contracted to facilitate their child's weight loss by carrying out reinforcement and stimulus control techniques, completing weekly charts and graphs, and encouraging their child to exercise. The response-cost only group parents did not contract to reinforce their child's performance. The response-cost program applied to both experimental groups was conducted in weekly meetings in which parents lost previously deposited sums of money. Twenty-five per cent was deducted for missing the weekly meeting, 25% for failing to fill out charts and graphs, and 50% if their child failed to meet her specified weekly weight-loss goal. At the end of the 12-week treatment period, both experimental groups had lost significantly more weight than the control group. After an eight-week, no-contact follow-up, some of the lost weight was regained. The response-cost plus reinforcement group was still significantly below the controls. The response-cost group just missed significance. A 31-week, no-contact follow-up failed to show a treatment effect, but did show a trend towards slower weight gain by the response-cost plus reinforcement group.