Research has shown that engaging students in peer feedback can help students revise documents and improve their writing skills. But the mechanistic pathways by which skills develop have remained untested: Does receiving and providing feedback lead to learning because it produces more extensive revision behavior or is such immediate implementation of feedback unnecessary? These pathways were tested through analyses of the relationships between feedback provided and received, feedback implemented and overall revisions, and improved writing quality in a new article. Overall, the number of revisions predicted growth in writing ability, and both amount of received and provided feedback were associated with being more likely to make revisions. However, providing feedback was also directly related to growth in writing ability.