Freshman composition students at a community college in Maryland participated in a study to determine whether the talk-write method of writing instruction would have a more positive effect on freshman writing than would more conventional methods of instruction. Four classes totaling 58 students were randomly divided into an experimental and a control group. The experimental group used the talk-write method, a prewriting activity in which pairs of students talk over their writing plans with one another, while the control group received instruction in grammar, punctuation, and outlining. All students then wrote narrative, descriptive, expository, and argumentative essays, which were independently evaluated to find differences in the mean scores for each group. The papers were judged on 12 measures falling into the areas of general merit, mechanics, and total evaluation. The results revealed significant differences favoring the experimental group. The talk-write method produced statistically significant positive effects on students' performance in writing for eight out of the twelve measures considered, with the area of general merit benefiting the most. (HTH)