Characteristics of the attack‐target in laboratory tests appear to greatly influence the tendency for aggression to be initiated, maintained, and stopped. To address this question, accessibility, target movement, and vocalization among pigeon conspecifics were investigated in two different aggression tests: 1) paired aggression (PA); and 2) schedule‐induced aggression (SI). In the PA test, dominant and subordinate roles formed quickly, and soon aggression episodes started to decline. In part, the subordinate's yielding behavior appeared to discourage attack. The significance of the target bird's behavior on attack frequency was borne out by the finding in the SI test that pigeons attacked a shielded, live, active target more often than a passive one. Furthermore, using a specially constructed stuffed pigeon in the SI test, results indicated that programmed, combined aggressive vocalization and body movement evoked most attacks with the VT schedule, vocalization alone evoked the next highest number of attacks, body movement alone the third most, and combined silence and immobility the least attacks. These findings provide a basis for explaining some past reported results associated with target features and a means for selecting other target features for future study.