Four procedures to evaluate teaching (by students, peers, video-recordings, and direct measurements of student learning) and three uses of the evaluation results (improving teaching, personnel decisions, course handbooks) are reviewed in the light of empirical evidence. Special emphasis is placed on the timing and validity of student ratings and the instruments used. Since none of the procedures appear sufficient in and by itself, a multiple indicator approach, especially for personnel decisions, would seem to be the most defensible one.While it is essential to take evidence of teaching effectiveness into account in considerations for tenure and promotion, faculty must also be given opportunities to become professionals as teachers. Higher education units, designed primarily for this purpose, appear to be effective as judged by their clients (the faculty they have served), but have failed to make an impact on the faculty as a whole. What is requited now is an institutional commitment to quality instruction, i.e. a departmental policy on the evaluation of teaching and faculty development.