Autonomic activity was continuously recorded while primary psychopaths (P), secondary psychopaths (S), and nonpsychopaths (NP) listened to a series of repetitive tones and solved simple arithmetic problems. During rest, Group P had lower levels of skin conductance and autonomic variability than did Group NP, but the difference in autonomic variability disappeared during the arithmetic period. Group P was generally less autonomically responsive to the tones, particularly when GSR and cardiac responses were expressed in terms of individual differences in the range of autonomic activity possible. While GSR habituated at the same rate for all groups, cardiac deceleration and, to a lesser extent, digital vasoconstriction habituated more slowly for Group P than for Group NP. Moreover, Group P gave a relatively small cardiac deceleratory response to a novel tone presented after the repetitive tones. The results were interpreted in terms of autonomic, cortical, and sensory coupling. Several investigators (Eysenck, 1964; Hare, of nonpsychopathic individuals. One reason 196Sb; Lykken, 1957; Quay, 196S; Schach-for this situation is that relatively little releter & Latane, 1964; Tong & Murphy, 1960; vant research has actually been carried out. Walton, 1960) have suggested that certain Moreover, the results of this research, though features of psychopathic behavior may be suggestive, have not always been consistent, related to some form of autonomic (and in For example, the psychopath's resting level some cases cortical) dysfunction. In most of autonomic activity has been found to be cases, it is assumed that the psychopath has normal in some cases and below normal in a hypoactive autonomic nervous system others. Similarly, he has been found to be (ANS) with the result that he is deficient in autonomically hyperresponsive, hyporesponthose autonomic correlates of anxiety neces-sive, and normally responsive to a variety of sary for the development and maintenance of neutral and stressful stimuli, certain patterns of behavior. Although several Methodological differences between studies studies are consistent with this assumption, it may account for part of the discrepant findis nevertheless true that very little is really ings. However, a more important reason may known about the ways in which the function-be the fact that not all investigators have used ing of the psychopath's ANS differs from that the same or comparable criteria for the selec-,_.. , tJ t.T,i.vTT ui. ti on °f psychopathic subjects (5s). In many iThis research was supported by