ALL SUBJECTS WERE BORN IN ST. LOUIS, MO., BETWEEN 1930 AND 1934, ATTENDED ST. LOUIS PUBLIC SCHOOLS FOR 6 YEARS OR MORE, LIVED IN ST. LOUIS AS ADULTS, AND HAD IQ (INTELLIGENCE QUOTIENT) SCORES OF AT LEAST 85. INTERVIEWS WERE CONDUCTED BETWEEN JUNE 1965 AND SEPTEMBER 1966. CORRELATIONS AMONG THE FIVE LIFE EVENTS WERE DETERMINED, AND CAUSAL RELATIONSHIPS WERE EXPLORED. ALL FIVE EVENTS ARE INTERCORRELATED. THE EVENTS ALSO ARE PREDICTED BY EARLIER EXPERIENCES -- ARREST, MARIJUANA USE, PRECOCIOUS SEXUAL ACTIVITY, AND DRINKING. SUCH NESTING OF PROBLEM BEHAVIOR IS A COMMON FINDING IN LIFE HISTORY RESEARCH. AN ACTUARIAL ANALYSIS OF DATA ON THE AGE OF ONSET FOR EACH LIFE EVENT, WHICH TAKES INTO ACCOUNT THE ORDER OF EVENTS, THE AGE SPAN OF EXPOSURE TO RISK, AND THE PRECURSORS COMMON TO POSSIBLE CAUSES AND OUTCOMES, INDICATES THAT HEROIN USE, INDEPENDENT OF PAST BEHAVIORS THAT LED TO HEROIN USE, CLEARLY PREDICTS LATER DEVIANT BEHAVIOR. DROPPING OUT OF SCHOOL SEEMS TO INCREASE THE RISKS OF LATER PROBLEMS, CLEARLY SO FOR ALCOHOL PROBLEMS AND POSSIBLY SO FOR OTHER PROBLEMS. ALCOHOL PROBLEMS MAY INCREASE THE RISK OF DIVORCE. THE MOST ELABORATE CAUSAL CHAIN SUGGESTED BY THE DATA IS THAT DROPPING OUT OF SCHOOL PREDICTS HEROIN USE, WHICH IN TURN PREDICTS ALCOHOL PROBLEMS, WHICH THEN PREDICT DIVORCE. TWO SHORTER CHAINS ARE ALSO SUGGESTED: DROPOUT, HEROIN USE, AND INCARCERATION/DIVORCE (WITHOUT ALCOHOL PROBLEMS INTERVENING); AND DROPOUT, ALCOHOL PROBLEMS, AND DIVORCE (WITHOUT HEROIN USE INTERVENING). INCARCERATION AND DIVORCE APPEAR ONLY AS CONSEQUENCES, NEVER AS PREDICTORS. THE FINDINGS SUGGEST THAT THE SOCIAL PROBLEMS OF YOUNG URBAN BLACK MALES COULD BE REDUCED BY FACILITATING THEIR GRADUATION FROM HIGH SCHOOL, BY DISCOURAGING HEROIN USE, AND BY DISCOURAGING HEAVY DRINKING. THE MOST CERTAIN BENEFITS WOULD COME FROM REDUCING EXPOSURE TO HEROIN. SUPPORTING DATA, A LIST OF REFERENCES, AND COMMENTS ON THE STUDY ARE INCLUDED. (LKM)