THE PRIMARY GOAL OF THE NOW-DEFUNCT KANSAS CITY PROGRAM WAS TO DEAL POSITIVELY WITH OFFICERS WHO RECEIVED MANY CITIZEN COMPLAINTS OR WHO FILED MANY 'INTERFERING WITH AN OFFICER' CHARGES. THESE OFFICERS WERE INVITED TO APPEAR BEFORE A PANEL OF THEIR PEERS TO DISCUSS THE ENCOUNTERS LEADING TO SUCH COMPLAINTS AND CHARGES. PANEL MEMBERS USED PEER PRESSURE AND SUPPORT TO ALTER THE OFFICERS' BEHAVIOR IN CONFRONTATIONS WITH CITIZENS. TO TEST THE EFFECTS OF THE PROGRAM, OFFICERS WHO APPEARED BEFORE PANELS WERE COMPARED WITH OFFICERS WITH SIMILAR HISTORIES WHO HAD NOT APPEARED BEFORE PANELS. THE BEHAVIORS, ATTITUDES, AND PEER EVALUATION RATINGS OF BOTH GROUPS WERE ANALYZED BEFORE AND AFTER THE PANEL APPEARANCES OF THE EXPERIMENTAL GROUP. DIVISIONWIDE EFFECTS WERE ASSESSED BY COMPARING THE EXPERIMENTAL DIVISION WITH OTHER PATROL DIVISIONS. NO SIGNIFICANT DIFFERENCES WERE FOUND BETWEEN EXPERIMENTAL AND CONTROL OFFICERS OR AMONG DIVISIONS. PROBLEMS THAT MAY HAVE LIMITED THE CONCLUSIVENESS OF THE EVALUATION ARE CITED. ONE SUCH PROBLEM WAS THAT EVALUATORS WERE DENIED ACCESS TO PANEL MEETINGS AND TRAINING SESSIONS. A BIBLIOGRAPHY AND SUPPORTING DATA ARE INCLUDED.