Historically, health-related quality of life research has been carried out at a descriptive level, providing rich data regarding the impact of disease and treatment on the physical, functional, psychologic, and social health of varying patient populations. More recently, there has been growing interest in incorporating psychosocial or "quality of life" outcome measures into clinical trials of medical interventions, particularly in the chronic diseases. This article reviews a number of the central methodologic issues surrounding the development or selection of quality of life measures appropriate for use in clinical trials. Topics discussed include the following: 1. Who should assess quality of life? 2. What should be assessed? 3. What means of data collection should be used, e.g., interviews, questionnaires or diaries? 4. Should the focus of the quality of life measures be generic or disease specific? 5. What are some of the considerations for response scales and for the time frame of questions addressing the quality of life? 6. What psychometric properties should be considered in evaluation of a questionnaire assessing quality of life? Additional attention is directed toward research design and implementation issues in clinical trial-based quality of life studies.