"PURE word-blindness" has long been a subject attractive to writers devoted to neurologically based language disorders. Authors principally concern themselves with clinical descriptions and speculations relevant to possible pathophysiological mechanisms. The literature leaves uncertainty regarding the natural history or prognosis of dyslexia. This observation might be broadly applied to the field of aphasia in its entirety. The paucity of works and controlled studies dealing directly with recovery from aphasic language disorders may simply reflect silent discouragement accrued from long experience in this field. Neurologists generally agree that variable but unpredictable potential for improvement exists, but documentation is scanty.1-5 Relatively "pure" cases of word-blindness are infrequently recognized. Inasmuch as I have had the opportunity of following two cases of agnosic dyslexia for two and four years respectively, an attempt was made to demonstrate any change occurring in these interesting patients. Early and current performances with interval history have been