Conscientious instructors at all levels are constantly making decisions about how to improve their teaching. This is a natural, intuitive part of being a good instructor, but even faculty who value student learning and work hard to improve their teaching can fail to make the connection between what they do in the classroom and the continuous improvement philosophy of outcomes assessment. The connection can be missed because understanding course-level outcomes assessment requires understanding evaluation policies and structures that differ markedly from those employed for unit-level program accreditation. The purposes of this paper are two-fold: first, to address policy issues that impede adoption of systematic, course-level outcomes assessment for improving student learning; and second, to describe a design structure for applying outcomes assessment that addresses factors under control of the instructor that affect students' learning. The paper concludes with observations on the crossed purposes of collecting information for judging faculty merit in the annual review process as opposed to collecting information for course improvement in the outcomes assessment process.