A method of measuring confrontation-induced behavioural changes in common marmosets (Callithrix jacchus) together with automated monitoring of locomotor activity has been developed as a possible model of anxiety. Recording both affiliative and agonistic behaviours between male/female pairs of marmosets and using diazepam as a reference drug, it has been possible to define a profile of behavioural changes which could be regarded as representing an anxiolytic response. Unfamiliar male/female pairs of marmosets were brought into close (non-contact) proximity in a controlled environment, in which their locomotor activity was recorded automatically. Simultaneously, their interactive behaviour was assessed by an independent observer via closed-circuit television. The following behaviours were analysed: aggressive postures, allogrooming, scratching, anxiety-related behaviours, social contact and self-grooming. Administration of diazepam at 1 and 3.5 mg/kg PO induced a significant (compared to control) reduction in scratching, aggressive behaviours, anxiety-related behaviours and an increase in allogrooming without affecting locomotor activity during confrontation. Differing responses dependant on gender were not found, nor did gender influence the effect of treatment on behaviour. Habituation to repeated confrontation did not occur. The results from this study demonstrate that this method can be used to measure anxiolytic activity in an objective manner. (C) 1997 Elsevier Science Inc.