This paper reports on an investigation of the evaluation of teaching and research at Stanford University in the School of Humanities and Sciences. It was found that professional colleagues had far more influence on organizational rewards and penalties than did administrative superiors. Evaluations of research had far more influence than did evaluations of teaching. Since the faculty expended considerable effort on teaching, many wanted a better balance between effort and reward. For teaching, students were influential evaluators, since teaching performances were more visible to them. For research, colleagues in other institutions were influential evaluators, since the visibility of published research made the boundaries of the University permeable. Overall, for both teaching and research, department colleagues and department heads were the most influential evaluators. Faculty members' satisfaction with the evaluation process was positively associated with the perceived dependence of their discipline on theory. Greater dependence on theory in a discipline was associated with more perceived agreement among evaluators, and more agreement among evaluators was positively related to satisfaction with the evaluation system. (Author/HS)