Lateralized words varying in size were postexposurally masked using checkerboards of 2, 4, or 8 cycles per degree. In errors, main effects were found for size, visual field, and spatial frequency. Although the expected right visual field advantage in recognition was found, and the differential effectiveness of the spatial-frequency masks was supported, there was no visual field × frequency interaction. In reaction time, a size × visual field interaction was found, but it was opposite to the effect typically claimed for size as an input variable. It is concluded that views of hemispheric differences that stress process type are likely to have more explanatory value than those stressing spatial frequency.