Although widely used, student evaluations of teaching do not address several factors that should be considered in evaluating teaching performance such as new course preparations, teaching larger classes, and inconvenient class times. Consequently, the incentive exists to avoid certain teaching assignments to achieve high SET scores while minimizing workload. This hinders curriculum innovation and development and potentially creates dysfunction in the college. The authors demonstrate how conjoint analysis can be applied to create a model of teaching evaluation that simultaneously considers many aspects of teaching performance and increases fairness in the appraisal process. Findings from a pilot implementation are discussed.