Abstract
The investigation of two patients with acquired dyslexia is reported. Both showed abnormally strong effects of word length on reading responses and occasionally named aloud the letters of words. By these criteria, they could be categorised as letter-by-letter readers. However, it was found that the patients had different functional deficits from one another and that the abnormally strong word-length effects were not necessarily a consequence of the same compensatory reading strategy. Further, the strategy did not necessarily involve the serial identification of letters. We suggest that the categorisation of these and other similar patients as letter-by-letter readers is irrelevant and misleading, not only because it does not indicate the functional impairment involved, but also because it fails to describe the strategies the patients are adopting to read and may lead to the inappropriate application of therapy.