It has been suggested that the differences observed for dyslexic readers compared to normal readers on tasks measuring visual sensitivity may simply be the result of differences between the two groups in general cognitive ability and/or attentional engagement. One common way to accommodate this proposal is to match normal and dyslexic readers on IQ. However, an explicit test of this suggestion is to take normal and dyslexic readers who differ on IQ--where IQ would be expected to explain reading ability--and determine if visual sensitivity can still account for reading skill, even when IQ is taken into account. In this study we explored the relative contributions of nonverbal IQ, visual sensitivity as measured by sensitivity to the frequency doubling illusion, and phonological and irregular word reading to reading ability. Visual sensitivity explained a significant amount of variance in reading ability, over and above nonverbal IQ, accounting for 6% of the unique variance in reading ability. Moreover, visual sensitivity was related primarily to irregular word reading rather than to nonsense word decoding. This study demonstrates that low-level visual sensitivity plays an intrinsic role in reading aptitude, even when IQ differences between normal and dyslexic readers are contrived to maximize the contribution of IQ to reading skill. These results challenge the suggestion that impaired visual sensitivity may be epiphenomenal to poor reading skills.