A model for assuring quality in the development of course objectives and classroom and exit examinations is presented. The model was based on a pilot study with 131 faculty at the University of Central Florida. It was found that 91% of teaching faculty create 100% of the tests they use to evaluate student performance. The faculty seemed to use course descriptions fairly regularly to develop course objectives. Faculty do not use a taxonomy in developing teacher-made test. However, few faculty obtain data on the reliability of their testing devices. According to the model, faculty members would have a good course description from which they would develop realistic and attainable course objectives. The course objectives would then be used to develop classroom and exit examinations, as well as course content. A taxonomy of testing, such as Bloom's, should then be used as a guide to constructing tests. Reliability data derived from test evaluation should be used to improve instruction and the measuring devise itself. Since tests determine whether the student has mastered educational objectives, use of this model would be a part of a quality assurance program. The model would also provide one objective and measurable input to the complex process of faculty evaluation. (SW)