THE STUDY PLAN INVOLVED FORMING A GENERAL TAXONOMY OF CRIMINALITY, BREAKING THIS INTO SMALLER UNITS, AND SUBJECTING THE UNITS TO GENETIC ANALYSIS. THE INVESTIGATION INVOLVED ALMOST 100 LIFE HISTORY VARIABLES TAKEN FROM INTERVIEWS WITH 54 FEMALE PROBATIONERS AND THEIR SIBLINGS, PLUS 14 PSYCHOLOGICAL TEST SCORES FROM EACH OF THESE SUBJECTS. AN INTERCORRELATION MATRIX WAS FORMED OF ALL THE VARIABLES, AND FACTOR SCALES (COMPOSED OF VARIABLES WHICH HAD THE HIGHEST ABSOLUTE LOADING ON THE FACTOR IN QUESTION) WERE DEVELOPED. THE SAMPLE WAS BROKEN DOWN INTO 5 GROUPS ON THE BASIS OF SEX AND CRIMINALITY, AND GROUP MEANS AND STANDARD DEVIATIONS ON ALL 14 PSYCHOLOGICAL TEST SCORES, FACTOR SCALES SCORES, AND CERTAIN INTERVIEW VARIABLES (ALCOHOLISM, DRUG EXPERIMENTATION, NEUROTICISM, SUICIDE ATTEMPTS, AND NUMBER OF JUVENILE OFFENSES) WERE DERIVED AND COMPARED. RESULTS SHOW THAT THOSE PERSONS WHO HAVE HAD SIGNIFICANT CONTACT WITH THE CRIMINAL JUSTICE SYSTEM HAD A LONG HISTORY OF PROBLEMS, INCLUDING EARLY PARENTAL SEPARATION, CRIMINALITY IN FATHERS, AND ALCOHOLISM IN BOTH PARENTS. THESE PROBLEMS HAVE BEEN CARRIED OVER INTO ADULT LIFE WITH A PATTERN OF DRUG ADDICTION, SUICIDE ATTEMPTS, AND LOW INCOME. IN ADDITION, OFFENDERS SHOW TENDENCIES TOWARD HOSTILITY, AGGRESSIVENESS, AND SELF-ALIENATION. THE NEXT STEP, GENETIC ANALYSIS, WEIGHED THE IMPORTANCE OF GENETIC FACTORS AND EXAMINED THEIR INTERACTION INSIDE A POLYGENIC MODEL WITH A THRESHOLD. THE MODEL ASSUMES THAT THERE ARE SMALL GENETIC AND ENVIRONMENTAL EFFECTS WHICH CAN BE TERMED LIABILITY. THE RESULTING CORRELATION IS 0.50 FOR FEMALES AND 0.12 FOR MALES, INDICATING THAT THERE IS A DIFFERENCE BETWEEN THE SEXES, ESPECIALLY IN THE ROLE PLAYED BY NONTRANSMISSIBLE ENVIRONMENTAL INFLUENCES ON THE TENDENCY TO CRIMINALITY. THE GENOTYPIC CORRELATION, WHICH ESTIMATES THE EXTENT TO WHICH THE LIABILITY IN THE TWO SEXES DEPENDS UPON THE SAME GENES, IS HIGH, BUT THE FACT THAT THE ESTIMATES IN RELATIVES ARE NOT THE SAME SHOWS THAT THE GENETIC CORRELATION FOR THE SEXES IS NOT EQUAL. THE FINDINGS INDICATE THAT CRIMINALITY AND PSYCHOPATHY ARE FAR MORE COMPLEX THAN PREVIOUSLY KNOWN, AND THAT VARIABLES USED HERE PROBABLY HAVE MANY SUBTYPES. TABULAR DATA AND REFERENCES ARE PROVIDED. (DAG)