Using the Global Assessment of Functioning (GAF) reminds me of the question my high school friends and I used to ask each other: "Do you walk to school, or do you carry your lunch?" This question, which invariably produced gales of laughter, was intended to confuse, because it cannot be answered with a simple "Yes" or "No"—both are things you might do on your way to school, but they are separate and not necessarily related.

So it is with the GAF, which combines several dimensions of psychopathology on a single 100-point scale. Specifically, it combines behavioral anchors for a mix of psychological functioning (mostly symptoms) and social, occupational, and school functioning (mostly interpersonal interaction and role performance). The behavioral measures for psychological functioning are joined to the measures for social, occupational, or school functioning by "or" statements—for example, "serious symptoms or any serious impairment in functioning." The GAF rating can be made easily only when symptoms and social functioning are in the same range of impairment.

However, even when the areas of functioning are not in the same range, one can learn to use the GAF. The rules direct the rater to assess the different manifestations of psychopathology and to make a rating based on the lowest level of functioning. Practically speaking, that might mean giving someone a GAF rating as low as 41 if he or she had "serious symptoms" but also had "some difficulty in social functioning" that would otherwise suggest a rating of 70. Some people with "serious" suicidal ideation can function with only "some" difficulty in social, occupational, and school situations, and it is a challenge to make ratings for these patients. Even when the assessment can be made reliably by assigning the lower rating, the clinician or investigator senses that some important information has been lost.

These concerns were expressed by the multiaxial issues workgroup that was formed to consider axis V, among other multiaxial issues, for the fourth revision of the DSM. The workgroup discussed the pros and cons of the GAF in the context of other measures of social functioning (1), and, despite reservations, recommended that the GAF be used to assess axis V. The group also suggested further testing of a separate scale, the Social and Occupational Functioning Assessment Scale (SOFAS), which appears in Appendix B of DSM-IV. The SOFAS pulls out from the GAF the measures of social, occupational, and school functioning to be rated separately, using the same anchors and rating scale. Dividing the GAF into its components has some conceptual appeal for overcoming some of the instrument's limitations. Interest in the SOFAS has been modest, and studies of its reliability and validity produced mixed results (2). Other multidimensional scales exist for making multiaxial assessments (3), and once again the process of DSM revision is examining the multiaxial approach to clinical assessment.

This issue of Psychiatric Services contains seven papers on the GAF. Each was submitted independently over the past year—indirect testimony to continued interest in the GAF—and we decided to organize them into a special section. As readers will see, results about the GAF's utility continue to be mixed. It is hard to deny the appeal of a single measure that is easy to use and that attempts to convey much information. The GAF also introduces elements of social, occupational, and school functioning into routine clinical assessment. Although these domains of functioning are at the heart of the impairments associated with mental disorders, they are not well captured by the diagnostic criteria within the DSM.

Many U.S. states are developing approaches for routinely assessing outcomes as a part of performance measurement (4). Four papers in the special section focus on the reliability and validity of the GAF in the Department of Veterans Affairs (two papers by Greenberg and Rosenheck) and in Sweden (two papers by Söderberg and colleagues). An article by Burlingame and colleagues reviews a state hospital's approach to outcome assessment, including the GAF, and a paper by Mulder and colleagues describes how the GAF was used as one of the determinants of use of a mobile psychiatric emergency service. A brief report by Shedlack and colleagues focuses on the limitations of the GAF for evaluating individuals with co-occurring mental retardation and mental illness. All these papers add to accumulating evidence of the advantages and disadvantages of the GAF.