PLASMA TESTOSTERONE LEVELS AND TESTOSTERONE PRODUCTION RATES OF THE SAMPLE POPULATION WERE CORRELATED WITH THEIR SCORES ON A BATTERY OF ANXIETY, DEPRESSION, AND HOSTILITY TESTS. THE FINDINGS INDICATE THAT A HIGHLY SIGNIFICANT RELATIONSHIP EXISTS BETWEEN TESTOSTERONE PRODUCTION RATES OF THE YOUNGER MEN AND A GROUP OF HOSTILITY AND AGGRESSION INDICATOR SCORES ON THE BUSS-DURKEE INVENTORY. AGE, RATHER THAN AGGRESSION, WAS THE PRINCIPAL DETERMINANT OF TESTOSTERONE PRODUCTION FOR THE SAMPLE OF OLDER MEN. THERE WERE NO SIGNIFICANT RELATIONSHIPS BETWEEN ANY OF THE AGGRESSION OR HOSTILITY INDICATORS FOR THIS OLDER GROUP AND TESTOSTERONE PRODUCTION, DESPITE THE FACT THAT PSYCHOLOGICAL TEST SCORES OF AGGRESSION AND HOSTILITY DIFFERED BUT LITTLE BETWEEN THE OLDER AND YOUNGER SUBJECTS. THE DIFFERENCES IN THE AGGRESSION-TESTOSTERONE RELATIONSHIPS BETWEEN GROUPS ARE POSSIBLY DUE TO QUESTIONABLE VALIDITY IN SELF-REPORTING OR TO A HIGHER DEGREE OF COPING WITH AGGRESSIVE FEELING EVIDENCED IN THE OLDER MALES. A MULTIVARIATE REGRESSION EQUATION WAS DERIVED UTILIZING TESTOSTERONE PRODUCTION RATES AS THE DEPENDENT VARIABLE AND FOUR PSYCHOLOGICAL MEASURES OF AGGRESSION AND HOSTILITY AS INDEPENDENT VARIABLES. THE RESULTING MULTIPLE CORRELATION COEFFICIENT ACCOUNTED FOR 82 PERCENT OF THE VARIANCE IN THE PRODUCTION RATE OF TESTOSTERONE FOR THE YOUNGER GROUP. THIS EQUATION WAS NOT VALID FOR THE OLDER MEN. THE GROUP OF PSYCHIATRIC PATIENTS WAS TOO SMALL FOR STATISTICAL COMPARISON OF MEANS. TABULAR DATA AND REFERENCES ARE PROVIDED. (AUTHOR ABSTRACT MODIFIED--DAG)