After attention has been oriented to a location, inhibition mechanisms prevent the return of attention shortly afterward. This inhibition can be associated with an object in such a way that after cuing attention to the object, inhibition can move with the object to a new location. Recent research has noted that the object-based inhibition of return effect in moving displays is much smaller than the effect observed in static displays, and hence may be of little functional utility. However, we demonstrate that, on the contrary, the large effects observed in static displays are produced precisely because of the existence of object-based frames, which can be additive with location-based frames of reference.Many visual environments are highly complex, and hence attention systems have evolved to select particular perceptual inputs to guide action. One of the main problems to be solved when organisms interact with complex environments is how objects relevant to behavior are found. Therefore, efficient mechanisms of visual search are crucial. One ofthe main features ofan efficient search system is that it be able to move to novel locations until a relevant object is discovered. That is, attention must not repeatedly return to recently examined loci since target stimuli may never be found. The importance of this point can easily be appreciated when considering an animal foraging for food.Posner and Cohen (1984) discovered a mechanism that would facilitate such efficient search. They showed that after attention had been automatically oriented to a spatial location via a peripheral onset cue, processing of stimuli at the cued location was impaired. They argued that this impairment reflected an inhibitory mechanism preventing the return of attention to the previously attended location. Posner and Cohen called this mechanism inhibition ofreturn (lOR) and said that its function was to facilitate search of novel loci.lOR has now been observed in a wide range of tasks. It is obtained when subjects detect the onset of targets with either keypress or saccades (see, e.g., Abrams & Dobkin, 1994), when they identify colors (see, e.g., Lupiafiez, Milan, Tornay, Madrid, & Tudela, 1997), letters (Lupiafiez, Milliken, Solano, Weaver, & Tipper, 1998), or auditory frequency (Mondor, Breau, & Milliken, 1998); and in cross-modal studies, in which the cue is presented in the visual modality and the target is presented in the auditory modality (Reuter-Lorenz, Jha, & Rosenquist, 1996). Of most pertinence to the aims of this article, the frames of reference that mediate lOR have also been investigated.It has been demonstrated (e.g., Maylor, 1985;Posner & Cohen, 1984) that lOR is based on environmental rather than retinal coordinates. Thus, when eye movements were This research has been supported by BBSRC Grant S06761 to S.P.T. Correspondence should be addressed to S. P.Tipper, School of Psychology, University of Wales, Bangor, Gwynedd, Wales LL57 2DG (e-mail: pss060@bangor.ac.uk).made between the cue offset and target onset, inhibition was as...