Two experiments investigated the effects of differing perceptual organizations of reversible figures on McCollough aftereffects. Experiment 1 used colored checkerboard inducing stimuli and achromatic grating test stimuli. While some subjects tended to organize the checkerboards into rows and/or columns and others to organize them into obliques, these variations did not result in differences in aftereffect direction or magnitude. Experiment 2 induced an aftereffect with colored gratings and tested with checkerboards, gratings, and a reversible concentric octagon pattern. Perceptual organization had no effect on results for checkerboards, but was related to aftereffect strength for the octagon pattern. Indirect evidence suggests that, in the latter case, differences in aftereffect strength may have influenced the perceived organization, rather than vice versa. Finally, regardless of the specific organization perceived, spontaneous viewing of all test stimuli produced stronger aftereffects than were found when subjects reorganized the pattern. This may have resulted from a viewing strategy associated with reorganization, since similarly small aftereffects were found when subjects concentrated their attention on a single pattern element.The McCollough effect can be elicited by exposing subjects to chromatic grating patterns. The color aftereffects are contingent upon the orientation of achromatic test gratings and are roughly complementary to the adapting color (McCollough, 1965;Skowbo, Timney, Gentry, & Morant, 1975;Stromeyer, 1978).Most attempts to account for these aftereffects have proposed the adaptation of neural analyzers that are specific to both color and orientation (McCollough, 1965) and, in some cases, to other features, such as movement direction (Hepler, 1968), curvature (Riggs, 1973), and spatial frequency (Stromeyer, 1972). These explanations are based on relatively low-level, involuntary mechanisms. Such interpretations are enticing, because, in most cases, the effects are associated with retinal stimulus properties rather than the properties as perceived by the subject (Harris, 1970;Stromeyer, 1978 (1977) found strong color aftereffects when subjects organized their concentric-square test pattern into four triangular sections of stripes. But little or no aftereffect was observed when the test pattern was organized as concentric squares. These results have important implications for our understanding of McCollough aftereffects. They suggest that the effects may not be as inflexible or as dependent on the physical stimulus properties as neural analyzer explanations would predict. Instead, they argue that higher order cognitive factors must be included for a complete explanation.However, conclusions from the results of previous studies with reversible figures must be tentative for several reasons. The first has to do with the distinction between the empirical results obtained and the internal condition of the subjects that is inferred from them.Both Jenkins and Ross (1977) and Uhlarik et al. (1977) ...