The faculty of an intensive program of English as a second language for college-bound students, questioning the objectivity of faculty evaluations of non-native college applicants' written essays, assessed the existing evaluation process, reformed it, tested it, and planned for ongoing development. In the first stage, readers read and graded essays according to practices currently in use, approaching the essays holistically, without breakdown of elements, and students wrote on a choice of topics. The lack of precision in grading was examined. The second stage involved devising new evaluation criteria. Faculty at five levels of English instruction analyzed instructional program goals and objectives, developed criteria from them, and applied them to the previously-graded essays. In the third stage, the revised criteria were tested in a midterm examination and the essays assigned randomly to the faculty for grading. The new evaluation process resulted in individual grids of writing strengths and weaknesses to be presented to the students. The fourth stage, assessment of the new criteria's usefulness, showed that the faculty valued these elements of the process; reduced evaluation time, a more humanistic testing procedure, helpful faculty discussion, more professional logistics, and a spirit of cooperation. Negative comments focused on the format of the grid of student strengths and weaknesses, the time involved in recording grades, and the use made of the results. The fifth stage involves an ongoing examination of the criteria's usefulness and reliability. (MSE)