ABSTRACT  To insure the successful use of writing exercises in the first-year language course, students must be asked to write only what they can understand and speak, and written work must be evaluated with an objective grading design. A three-point grading design used for evaluating single sentences and a five-point grading design appropriate for student-produced dialogues allow for that objectivity. Measuring the written work of students against predetermined criteria enables the instructor to examine analytically the contributions of students who are developing at individual rates of progress without making favorable or unfavorable comparisons among students. The objectivity and reliability of the grading design was tested in an experiment in which three instructors each evaluated thirty themes. The correlation between the three sets of grades was determined to be 0.93.