A 10-ITEM SCALE WAS EMPLOYED TO DIVIDE 68 MALE INMATES INTO HIGH (GROUP H: N=15), MEDIUM (GROUP M: N=33), AND LOW (GROUP L: N=20) ANTISOCIAL GROUPS. ELECTRODERMAL ACTIVITY WAS MEASURED IN RESPONSE TO A 1000-HZ, 75-DB TONE OF 2 SECONDS DURATION, PRESENTED AT RANDOMLY ORDERED INTERVALS OF 15, 20, AND 25 SECONDS. HALF-RECOVERY TIME WAS MEASURED AS THE TIME TAKEN FROM PEAK-RESPONSE AMPLITUDE TO HALF-RESPONSE AMPLITUDE; RECOVERY RATE WAS MEASURED BY DIVIDING ONE HALF OF THE RESPONSE AMPLITUDE BY THE HALF-RECOVERY TIME. ANALYSES OF VARIANCE PERFORMED ON THE RECOVERY DATA INDICATED SIGNIFICANT EFFECTS FOR BOTH HALF-RECOVERY TIME AND RECOVERY RATE. FURTHER ANALYSES OF THE DATA INDICATED THAT GROUPS H AND M DISPLAYED SIGNIFICANTLY SLOWER RECOVERY THAN DID GROUP L. TO DETERMINE HOW WELL THE RECOVERY MEASURES CLASSIFIED SUBJECTS, EACH GROUP WAS DIVIDED INTO THOSE ABOVE AND BELOW THE MEDIAN OF THE POOLED GROUPS. BOTH VARIABLES DIFFERENTIATED THE GROUPS AT A STATISTICALLY SIGNIFICANT LEVEL, ALTHOUGH THE DIFFERENTIATION WAS BEST FOR GROUP H. ALTHOUGH THE HYPOTHESIS IS APPARENTLY ESTABLISHED BY THE TEST, IT IS NOTED THAT BECAUSE EDR RECOVERY HAS BEEN SHOWN TO BE SITUATIONALLY DETERMINED, FURTHER WORK MUST BE AIMED AT REPLICATING THE PRESENT FINDINGS, USING MORE NOXIOUS STIMULATION AS PUNISHMENT FOR INAPPROPRIATE BEHAVIOR. TABLES OF DATA AND RESULTS ARE INCLUDED. (RCB)