In this study a method of teaching composition using individualized learning and peer grouping was developed, implemented, and evaluated. The model was used with 30 students in a tenth grade English class while a control group of 30 students received more traditional instruction. It was found that significantly greater gains were made by the experimental group in organization, critical thinking, and appropriateness, while the control group had greater growth in conventions. It was concluded that peer evaluations of compositions tended to be at least as effective as teacher correction and greatly reduced the need for out-of-class teachr time spent in evaluation. Peer evaluation also enabled students to complete more compositions while receiving more feedback on each writing. In the individualized phase of the model students progressed at their own rate in acquiring composition skills without repeating previous learning. (Author/DI)