The effectiveness and efficiency of two error correction procedures on word reading were compared. Three students with below average reading skills and one student with average reading skills were provided with weekly instruction on sets of 20 unknown words. Students' errors during instruction were followed by either word supply error correction (the researcher said the word and the student was asked to repeat it) or a multilearning channel error correction procedure, which included four components: (a) hear word-say word, (b) see word-say letters (step repeated once), (c) think-say letters (spell without seeing word; step repeated once), and (d) think-write letters (write letters without seeing word). Both conditions generally improved reading performance, although one procedure was not clearly superior to the other in increasing the frequency of words read correctly. The word supply procedure was approximately twice as time efficient. (Contains 2 tables.)