The author examined whether independent or peer revision produced a greater number of meaningful revisions between drafts. The study was a single-participant, alternating treatment design. The sample comprised 8 students in a predominately low-SES urban school in the United States. Students of mixed gender and background were in a self-contained, multilevel class designed for students with significant behavioral and emotional disabilities. Students wrote daily in both treatment conditions in a counterbalanced fashion. Results indicate that students made more surface-level revisions in the independent condition and produced more content-related revisions in the peer condition. In both conditions, the 2nd draft was more descriptive than the 1st, and students in the peer condition increased their use of complex vocabulary and organizational components. In addition, discussion during peer conferencing was positive, appropriate, and relevant to the topic. (Contains 1 note and 4 tables.)