The purpose of this study was to compare effectiveness of a word study phonics technique called word sorts on word recognition, pseudoword recognition, phonological awareness, and spelling between a group of children with mental retardation and a group of struggling young readers without identified disabilities. Word sort instruction involved categorizing words into groups based upon common sound and spelling patterns. As sessions progressed, both groups of children were successful at completing the sorts. A MANCOVA revealed that there were no significant differences between the groups on posttest measures of word recognition, pseudoword recognition, phonological awareness, and spelling while holding pretest performance on these variables constant. There was variability in performance within groups, especially within the group of children with mental retardation. Some children within the groups made more substantial gains from pretests to posttests than others. Children without disabilities made greater gains on a pseudoword naming task. Implications are discussed for how the word sort technique provides an analogy-based approach for helping children read words unknown to them.