Dyslexia is a term widely used to describe reading characterised by problems with the fluent and accurate letter or word recognition. Nevertheless, there is no consensus about the definition, origin, and diagnosis of dyslexia and the term is often used very differently by researchers and practitioners. In many cases, research findings are employed by clinicians in ways that are misleading and potentially counterproductive. The present study takes the form of an examination of participant samples included in studies of dyslexia (n = 800) over 20 years (2000-2019). The findings show that: (1) researchers use a wide range of inclusion and exclusion criteria; that (2) IQ-reading achievement discrepancy is the most common inclusion criterion for dyslexia samples; (3) studies typically compare dyslexic samples to normal controls but not to other poor readers; (4) dyslexia seems to be employed as a catch-all term for poor readers in general, not as a term to define a specific type of poor reader. Finally, (5) dyslexia studies are very rarely published in educational journals.