The percentage of universities using student ratings to evaluate teacher performance is steadily increasing, but this method can be abused if it is the sole criterion of evaluation. Recent surveys of faculty attitudes towards teaching evaluation indicate that most faculty members are in favor of systematic evaluation, prefer student ratings to other methods of evaluation, and believe that student ratings should be used as one of several sources of information in salary, promotion, and tenure decisions. A comprehensive, fair teaching evaluation form to be used for summative purposes should be: easily understood by students; related to student learning; satisfactory in terms of psychometric criteria; under the instructor's personal control; and applicable to many different types or styles of teaching. Student ratings, to be reliable, should be examined for internal consistency, inter-rater reliability, and test-retest reliability. Research on the validity of student ratings is reviewed, and sources of bias are identified. Alternatives to student ratings in the evaluation of teaching might be direct measures of student achievement and colleague evaluations. Considering these alternatives, and certain limitations of student evaluations, student ratings of teaching are sufficiently reliable and valid to be used only as one of several sources of information in administrative decisions relating to faculty salary, tenure, and promotion. References are appended. (DC)