This article examines the possibility of a "Teaching Assessment Exercise" and attempts to quantify teaching quality as part of performance management schemes for academics. The primary sources of data are identified as student evaluation of teaching (SET) and peer observation of teaching (POT). The conceptual and empirical issues in developing valid and reliable teaching quality indices from SET and POT are critically reviewed. The difficulties of using such data for academic performance management are discussed, focusing on the tensions between using such data both formatively for professional development and summatively for decisions about employment. (Contains 2 notes.)