JUVENILE REFERRAL SYSTEMS ARE PROGRAMS THAT COORDINATE EXISTING COMMUNITY RESOURCES WITH THE NEEDS OF THE POLICE AND THE COURTS. THE SUBJECTS IN THIS STUDY WERE 149 JUVENILES WHO WERE SENT TO THE MONROE COUNTY YOUTH SERVICES SYSTEM (NEW YORK) BY THE ROCHESTER POLICE DEPARTMENT AND 90 JUVENILES WHO WERE NOT REFERRED. THE REFERRED YOUTHS WERE DIVIDED INTO TWO GROUPS: 77 IN THE HIGH SERVICE GROUP AND 72 IN THE LOW SERVICE GROUP. THE STUDY COMPARED EIGHT BACKROUND VARIABLES AND MEASURED DIVERSION BASED ON WHETHER YOUTH'S FAMILY COURT PETITION WAS DISMISSED OR IF THE YOUTH RECEIVED A DETENTION SENTENCE. IT MEASURED RECIDIVISM BY DETERMINING WHETHER YOUTHS HAD FURTHER POLICE CONTACT IN A 6-MONTH FOLLOWUP PERIOD. FINDINGS SHOWED THAT, COMPARED TO THE LOW SERVICE GROUP, THE HIGH SERVICE YOUTHS HAD A HISTORY OF MORE PRIOR CONTACTS WITH THE POLICE, WERE MORE LIKELY TO BE MALES, AND WERE MORE LIKELY TO BE MINORITY GROUP MEMBERS. PETITIONED REFERRALS HAD THEIR CASES DISMISSED MORE FREQUENTLY THAN PETITIONED NONREFERRALS. HOWEVER, REFERRALS DID NOT DIFFER FROM NONREFERRALS WHEN COMPARED FOR RECEIVING A DETENTION SENTENCE. FEWER REFERRALS WOULD BE RECIDIVISTS IN THE FOLLOWUP PERIODS THAN THE NONREFERRALS. HOWEVER, NO RELATION EXISTED BETWEEN AMOUNT OF SERVICE AND RECIDIVISM. OVERALL, THE PROGRAM REACHED MULTIPLE OFFENDERS MORE FREQUENTLY AND FUNCTIONED MORE AS A TREATMENT FACILITY THAN AS A PREVENTIVE SERVICE FOR FIRST OFFENDERS, ALTHOUGH A PREVENTIVE EFFECT WAS OBSERVED FOR THE SECOND 3 MONTHS OF THE FOLLOWUP PERIOD. THE SERVICE THAT YOUTHS RECEIVED FOR THEIR FIRST OFFENSE MAY HAVE SERVED AS A REWARD AND INCREASED THEIR PROPENSITY FOR ILLEGAL BEHAVIOR. THE STUDY RECOMMENDED FURTHER EXAMINATION OF PARTICIPATING AGENCIES IN ORDER TO REFER THE APPROPRIATE YOUTH TO THE APPROPRIATE AGENCY. FOOTNOTES, REFERENCES, AND A TABLE ARE INCLUDED. (JLF)