A battery of tests designed to elicit reactions to a variety of nonpainful threat stimuli was used to study the effects of the 5HT1A agonists buspirone (5-20 mg/kg), and gepirone (5-20 mg/kg) on the defensive repertoire of wild Rattus rattus. These two compounds produced very similar patterns of results on the test battery, with gepirone generally more effective: Both compounds failed to interfere with either spontaneous motor activity or avoidance/flight to an approaching experimenter. However, both reduced defensive reactivity to proximal threat stimuli, increasing passive contacts with the experimenter in an inescapable situation and reducing "proximal" defensive reactions: jump/flinch reactions to dorsal contact, and, boxing, and biting to a number of threat stimuli. Defensive threat vocalizations and jump attacks were also reduced, but less consistently, as was the experimenter's rating of subject's defensiveness to being picked up. This pattern of results suggested specific "taming" effects of buspirone and, especially, gepirone on defensive reactions. In combination with findings indicating somewhat different (benzodiazepines) to very different (ethanol) profiles for other anxiolytics in the same test battery, these results suggest that the Defense Test Battery may be capable of providing behavioural differentiation among various classes of anxiolytics.