Rarely in educational research do we have access to data that allow us to empirically explore the relationships between the practices of individual teachers and the learning of their students. This report is one of those exceptional cases. In this report, we use data from Plainfield, New Jersey in which individual teachers' survey responses about America's Choice were linked to the test gains of the students who were taught by those teachers. By working closely with district administrators, we were able to link individual survey responses to the district's student achievement databases while retaining the confidentiality of both teachers and students. The result is an uncommon piece of evidence that empirically links teachers' implementation of America's Choice to student learning. The pattern from these results seems clear and persistent: the students of teachers who more deeply implemented the America's Choice model, particularly the writers workshop component of the design, learned more than did the students of teachers who had lower levels of implementation. Even after statistically controlling for the background characteristics of teachers and students and for students' prior test performance, teachers' implementation of America's Choice was associated with significantly higher learning gains for students.