Using a single-subjects-with-replicates design, this study investigated conference influence on first graders' knowledge about revision as well as revision activity. Sixteen children participated in group writing conferences with a teacher, in a natural classroom setting, every other week from February through June. Data from three baseline points and seven conference points were summarized. At conference data collection points, students wrote, conferred in groups with a teacher, were interviewed about potential revisions, and revised work in progress. At baseline points, the same events occurred, but there were no conferences. Two main variables were used to evaluate knowledge of the revision process: number of spots suggested for revision and average specificity of suggested changes. The main variable for actual revision activity was total number of revisions made. Final drafts were also rated for quality. Conferences did influence revision knowledge and revision activity for many children. However, the extent of conference influence was mediated by certain entry level student characteristics. Generally, the most positive effects occurred for students who began with the least amount of knowledge about revision, who were initially doing the least amount of revision, and who were initially writing pieces judged among the lowest in quality.