The study investigated the use of phonological encoding and chunking strategies (the skimming and clustering of highly familiar material into meaningful units) with congenitally deaf students at the middle and upper levels of an oralist school as well as hearing students at the second, third, fourth, fifth, and college grade levels. Ss were asked to identify all instances of a particular letter in three passages presented in either normal prose, scrambled words, or scrambled letters. Results indicated that deaf Ss made fewer errors than did hearing Ss (suggesting less efficient reading by deaf Ss), and that both hearing and deaf Ss showed evidence for phonological encoding. A replication of the study with congenitally deaf children from a total communicaton program found equal evidence for speech recoding with no difference in comparison to the deaf children trained under the oral approach. Results demonstrated that the deaf child's difficulties with reading cannot be attributed to a reliance on an inefficient visual (rather than phonological) coding strategy. (DB)