Analyses used data from the Cambridge Study in Delinquent Behavior, a prospective longitudinal survey of boys of working class background begun in 1961-62. All but 12 were white. Psychologists interviewed and tested the boys in their schools when they were aged about 8, 10, and 14. At age 21, 218 of the target group of 241 convicted delinquents and a randomly chosen sample of unconvicted youths were interviewed. At age 24, 85 of the target group of 113 were interviewed. These consisted of four subgroups: persisting recidivists (those with two or more convictions up to age 19 and at least one more in the next 5 years), temporary recidivists (those with two or more convictions up to age 19 and no more in the next 5 years), unconvicted youths from seriously deprived backgrounds (from large families, in poor housing, with convicted parents, and with families supported by State welfare), and a random sample of unconvicted youths. Parents also were interviewed. Teachers completed questionnaires when the boys were aged about 8, 10, 12, and 14. Youths who were convicted or who admitted large numbers of delinquent acts were identified as troublesome, daring, dishonest, and aggressive by their teachers, peers, and parents from an early age. Over and above the continuity of behavior, some factors had persistent effects. The most important of the earliest factors were economic deprivation, family criminality, parental mishandling, and school failure. The most important of the later factors were truancy, delinquent friends, anti-establishment attitudes, and an unstable job record. Based on the research findings, a speculative theory is proposed to explain the most common varieties of male delinquency. Twenty-one references are included.