THE COMMON MEASURE BY WHICH THE PROJECTS WERE EVALUATED FOR THEIR ABILITY TO AFFECT THE CRIMINAL BEHAVIOR OF THEIR CLIENTS WAS ARRESTS. A STANDARDIZED INTAKE FORM WAS DEVELOPED, IMPLEMENTED, AND MONITORED TO PROVIDE THE EVALUATION WITH ACCURATE IDENTIFICATIONS AND BACKGROUND INFORMATION. IDENTIFICATIONS WERE USED TO RETRIEVE THE ARREST HISTORIES OF THE CLIENTS FROM THE NEW YORK CITY POLICE DEPARTMENT AND MEASUREMENT OF SEVERITY OF CRIMINAL HISTORY PRIOR TO PROJECT ENTRY WAS DETERMINED. RESULTS INDICATE THAT DIFFERENCES AMONG PROJECTS DID NOT AFFECT THE ARREST RECIDIVISM RATES OF SIMILAR TYPES OF CLIENTS; THAT THE MAGNITUDE AND SEVERITY OF RECIDIVISM WAS HIGH AND AFFECTED BY AGE AND CRIMINAL HISTORY; AND THAT VIOLENT CRIME BEFORE PROJECT ENTRY WAS RELATED TO VIOLENT CRIME AFTER PROJECT ENTRY. IT WAS CONCLUDED THAT REHABILITATION BY THE PROJECTS WAS A FAILURE. FOR AN APPENDIX DETAILING EVALUATION DESIGN AND METHOD, SEE NCJ-36345.