THE EXPERIMENTAL SUBJECTS WERE 30 MALES AND 30 FEMALES WHOSE BEHAVIOR (ATTEMPTED SUICIDE, ALCOHOL ABUSE, DRUGS, VIOLENCE, THEFT, OTHER CRIMES, SEXUAL PROMISCUITY) HAD NECESSITATED HOSPITALIZATION, BUT WHO WERE NOT DIAGNOSED AS SUFFERING FROM ANY ORGANIC CONDITION, PSYCHOSIS, OR SPECIFIC NEUROTIC REACTION. THESE PATIENTS AND RANDOM CONTROLS COMPLETED A PERSONALITY TEST DESIGNED TO MEASURE NEUROTICISM AND EXTROVERSION-INTROVERSION. THE IMPULSIVENESS AND SOCIABILITY SECTIONS OF THE EXTROVERSION SCALE WERE ANALYZED SEPARATELY. THE PATIENTS SCORED HIGHER THAN CONTROLS ON NEUROTICISM, BUT MEAN SCORES ON EXTROVERSION AND ITS COMPONENT SECTIONS DID NOT DIFFER BETWEEN THE TWO GROUPS. HOWEVER, THERE WAS GREATER VARIANCE AMONG PATIENTS THAN AMONG CONTROLS ON THE EXTROVERSION SCALE AND ON ITS SOCIABILITY COMPONENT. DYSTHYMIC (E.G., ATTEMPTED SUICIDE) AND PSYCHOPATHIC (E.G., THEFT, VIOLENCE) BEHAVIORS WERE ASSOCIATED WITH PATIENTS AT BOTH EXTREMES OF THE EXTROVERSION AND SOCIABILITY SCALES. IT MAY BE THAT EITHER EXTREME OF EXTROVERSION AND SOCIABILITY COMBINED WITH A HIGH DEGREE OF NEUROTICISM PREDISPOSES AN INDIVIDUAL TO ABERRATED BEHAVIOR, BUT DOES NOT DETERMINE THE QUALITY OF THAT BEHAVIOR. SUPPORTING DATA, DETAILS OF THE ANALYSIS AND INTERPRETATION, AND A LIST OF REFERENCES ARE INCLUDED. (LKM)