A measure of satisfaction is essential to understanding clients' preferences and providing feedback to professionals and managers. However, as a single measure for policy purposes it has significant limitations. It is often unclear which components of satisfaction are critical and which can be affected by changes in the system or in providers' behavior. Multidimensional measures of satisfaction should be used, and specific components that are most important for policy decisions should receive the highest weights. However, policy makers should not rely solely on satisfaction as an indicator of the utility of a given intervention. Information about other factors, such as clinical and functional status, quality of life, and adherence to treatment regimens, is necessary to understand the utility of an intervention, and these other factors may deserve more emphasis.