The authors reviewed 21 studies assessing housing preferences of mental health consumers to examine current methods of assessment, to obtain information on the reliability and validity of the assessment instruments used, and to determine whether a better approach to assessing preferences might be found. The review found little data on the reliability or validity of the assessment instruments and revealed heavy reliance on the use of fixed-choice questions that limit consumer expression of preferences. Responses to such questions can be misleading because they may or may not reflect real-life constraints, such as preferring a roommate only because living alone is not affordable. However, the authors believe the factorial survey model, which uses vignettes to present different combinations and characteristics of living arrangements, may allow investigators to better understand consumers' true preferences. They propose creating and pilot testing an instrument using such vignettes.