THE DATA WERE DRAWN FROM THE CAMBRIDGE (ENGLAND) STUDY IN DELINQUENT DEVELOPMENT, WHICH FOLLOWED A NORMAL POPULATION OF LONDON SCHOOLBOYS FROM AGE 8 TO AGE 23. THE FINAL SAMPLE CONSISTED OF 397 MEN FROM AN ORIGINAL SAMPLE OF 411 BOYS FROM A WORKING CLASS URBAN AREA BORN BETWEEN 1951 AND 1954. OF THESE, 35 WERE PERSISTENT RECIDIVISTS, WHILE 26 WERE TEMPORARY RECIDIVISTS WHO HAD TWO OR MORE OFFENSES RECORDED BEFORE THEIR 19TH BIRTHDAYS BUT WHOSE CRIMINALITY DID NOT EXTEND INTO ADULTHOOD. BOTH GROUPS ARE COMPARED WITH TABLES PRESENTING STUDY DATA. THE PERSISTENT RECIDIVISTS HAD A GREATER NUMBER OF CONVICTIONS BEFORE THE 19TH BIRTHDAY AND HAD HIGHER SCORES FOR 'ANTISOCIALITY' FACTORS ON PERSONALITY TESTS COMPARED WITH TEMPORARY RECIDIVISTS. PERSISTENT RECIDIVISTS WERE MORE LIKELY TO HAVE HAD POOR FAMILY INCOME AND LARGE-SIZED FAMILIES. THE MOST EFFECTIVE PREDICTORS OF ADULT CRIMINALITY, HOWEVER, WERE NUMBER OF PREVIOUS CONVICTIONS, ESPECIALLY JUVENILE CONVICTIONS, AND FAMILY CRIMINALITY. DATA GOING BACK BEFORE AGE 10 REVEAL THAT THE PERSISTENT RECIDIVISTS HAD A STRONG HISTORY OF ANTISOCIAL SCHOOL BEHAVIOR. ASSESSMENTS OF 'ANTISOCIALITY' BASED ON INTERVIEWS AT AGE 18 PROVED TO BE AS PREDICTIVE OF RECONVICTIONS AS OFFICIAL CONVICTION HISTORIES OR SELF-REPORTED DELINQUENCY SCORES. IT IS CONCLUDED THAT SOCIAL HABITS ARE IMPORTANT INDICATORS IN THE PROGNOSIS OF A CRIMINAL CAREER. LATECOMERS TO CRIME, CONVICTED FOR THE FIRST TIME AFTER THE 19TH BIRTHDAY, DID NOT SHARE THE EARLY BACKGROUND ADVERSITIES OF THE OTHERS BUT SHOWED TROUBLESOME SCHOOL BEHAVIOR FROM AN EARLY AGE. NUMEROUS TABLES DISPLAY STATISTICAL DATA. REFERENCES ARE PROVIDED. (GLR)