This chapter examines the assessment and measurement of mental disorders. Researchers must distinguish between clinical prevalence (people who are treated for mental disorder) and true prevalence (the actual rate of disorder in a community, including those not in treatment). The measurement of mental illness must be conceptually valid; that is, there must be criteria that successfully distinguish cases of disorder from cases of non-disorder. In the past, researchers relied upon general symptom checklists, which identify a threshold above which an individual is considered disordered, but without specifying a particular disorder. An alternative to checklists is provided by the American Psychiatric Association's Diagnostic and Statistical Manual (DSM) of mental disorders, which provides sets of diagnostic criteria for specific disorders. The assumption behind the DSM is that mental disorders result from internal psychological dysfunctions (i.e., failures of proper functioning of mental processes), a presumption that Wakefield and Schmitz accept but demonstrate is often violated by the DSM's own criteria for mental disorder. Their critique of the DSM's approach to measurement is illustrated with several DSM diagnoses. In addition to thoroughly discussing the conceptual basis of the DSM, Wakefield and Schmitz provide examples of the attempts to use DSM-derived criteria to measure prevalence of mental disorder in the community. These examples demonstrate the recurrent problems with creating conceptually valid measures for use in psychiatric epidemiology. It is unclear whether these problems can be overcome or circumvented with methodological innovations. The student should consider why it is so difficult to determine who is mentally disordered, and to distinguish mental disorder from intense normal distress. Is a conceptually valid resolution of these problems possible?