ALSO INCLUDED WAS A CASE-BY-CASE ANALYSIS OF OFFICERS' PREDICTIONS OF THE FUTURE DELINQUENCY INVOLVEMENT OF BOYS WITH WHOM THEY HAD FIRST CONTACT OVER A SIX-MONTH PERIOD AND TESTS OF THE ACCURACY OF THESE PREDICTIONS OVER A SIX-YEAR FOLLOW-UP PERIOD. DATA WAS COLLECTED IN THREE CITIES DURING THE SPRING AND SUMMER OF 1968 AND 1970 USING INSTRUCTURED INTERVIEWS. QUESTIONNAIRE RESULTS INDICATED THAT POLICE IN WIDELY DIFFERENT SETTINGS BASE THEIR JUDGMENTS ON 'STREET' EXPERIENCE AND SIMILAR THEORIES OF DELINQUENCY CAUSATION LINKING SOCIAL CLASS BACKGROUND, PARENTAL NEGLECT, AND DELINQUENT BEHAVIOR. THEY SAW LOWER CLASS BOYS AS MORE LIKELY TO BE INVOLVED IN DELINQUENT CONDUCT THAN BOYS WITH OTHER FAMILY AND SOCIAL CLASS BACKGROUNDS. THESE VIEWS WERE CLOSELY MATCHED BY OFFICIAL CONTACTS, THOUGH PATTERNS VARIED IN EACH OF THE CITIES STUDIED. OFFICERS' PREDICTIONS FOR INDIVIDUAL BOYS PROVED TO BE INACCURATE, MEASURED BY OFFICIAL CONTACT RECORDS OVER A SIX-YEAR PERIOD. QUESTIONNAIRE EXCERPTS ARE APPENDED. REFERENCES ARE INCLUDED. (AUTHOR ABSTRACT MODIFIED)