This paper argues that the sort of evaluation and measurement needed to support language learning and teaching is significantly different from the sort of measurement that most language testing experts have traditionally been interested in. Particularly, the classroom teacher’s need for continuous evaluation and the necessity for teachers to help students develop the self-evaluation skills that they must have if they are to become lifelong learners make the criteria of classroom evaluation different from the concerns of most language testers. This paper tries to make the case for a different approach to classroom measurement by showing that the periodic measurement devices that most testing experts have focused on are not adequate in the classroom situation. It then makes specific suggestions for the development of continuous evaluation and suggests ways in which students can be taught self-assessment. It also claims that most language testers lack sufficient knowledge about the language classroom and the limitations under which teachers work to be able to simply tell them what they should be doing. Therefore, dialogue between testers and teachers classroom teachers is needed. This dialogue is essential because the central concerns of all testing (validity, reliability, and practicality) are needed for successful classroom evaluation.