In this paper we argue that attempts to distinguish between categories of "dyslexia" and "poor reader" or "reading disabled" are scientifically unsupportable, arbitrary and thus potentially discriminatory. We do not seek to veto scientific curiosity in examining underlying factors in reading disability, for seeking greater understanding of the relationship between visual symbols and spoken language is crucial. However, while stressing the potential of genetics and neuroscience for guiding assessment and educational practice at some stage in the future, we argue that there is a mistaken belief that current knowledge in these fields is sufficient to justify a category of dyslexia as a subset of those who encounter reading difficulties. The implications of this debate for large-scale intervention are outlined.