THE YOUNGSTERS, AGED 11 TO 17, HAD BEEN REFERRED FOR BEHAVIOR AND ACADEMIC PROBLEMS. THE VOLUNTEERS RANGED IN AGE FROM 17 THROUGH 65 AND IN EDUCATION FROM COMPLETION OF THE FOURTH GRADE THROUGH COMPLETION OF GRADUATE LEVEL STUDY. THE BUDDIES WERE PAID UP TO $144 A MONTH FOR MAKING WEEKLY CONTACT WITH EACH OF THEIR ASSIGNED YOUNGSTERS, SUBMITTING WEEKLY REPORTS, AND ATTENDING BIWEEKLY TRAINING SESSIONS. THREE INTERVENTION STRATEGIES WERE DESIGNED: (1) THE BUDDY WAS ACCEPTING OF EVERYTHING THE YOUNGSTER DID AND REWARDS DID NOT DEPEND UPON BEHAVIOR, (2) SOCIAL APPROVAL WAS CONTINGENT UPON ACCEPTABLE BEHAVIOR BUT MONETARY REWARDS WERE NOT, (3) BOTH SOCIAL AND MONETARY REWARDS WERE CONTINGENT UPON BEHAVIOR. CONTROLS WERE YOUNGSTERS REFERRED TO THE BUDDY SYSTEM BUT WHO COULD NOT BE ACCEPTED. SCHOOL ATTENDANCE IMPROVED WHEN EITHER THE SOCIAL APPROVAL AND SOCIAL-MONETARY APPROVAL SYSTEMS WERE USED, BUT NOT WHEN THE ACCEPTING INTERVENTION WAS USED. A COMBINATION OF SOCIAL APPROVAL AND MATERIAL CONTINGENCY WAS ALSO EFFECTIVE IN DECREASING FIGHTING, RETURNING HOME LATE, AND NOT DOING HOME CHORES. THE EXPERIMENT WAS DIVIDED INTO TWO 6-WEEK PERIODS. THE CONTINGENCY GROUPS MAINTAINED THEIR IMPROVED BEHAVIOR THROUGHOUT THE ENTIRE PERIOD. THE ACCEPTING GROUP GRADUALLY BEGAN TO SHOW IMPROVED ATTENDANCE DURING THE SECOND 6 WEEKS. SCHOOL GRADES DID NOT IMPROVE FOR ANY GROUP. IT IS SUGGESTED THAT ACADEMIC PERFORMANCE MUST BE TARGETED AND TEACHERS BE USED AS THE CHANGE AGENTS. REFERENCES ARE APPENDED. (GLR)