In this first article in a two-article series, I diagnose' a problem with DSM-IV, specifically, the overinclusiveness of its diagnostic criteria. Using the harmful dysfunction analysis of the concept of disorder (Wakefield, 1992a, American Psychologist, 47, 373-388) as a framework, I argue that DSM-IV criteria for many diagnostic categories fail to satisfy the analysis' dysfunction' requirement, that is, the criteria do not distinguish harmful conditions due to internal dysfunctions from harmful conditions that are nondisordered 'problems in living'. The overinclusiveness problem, I suggest, can be partly dealt with by giving up purely symptomatic criteria and contextualizing diagnosis to take into account the relationship between triggering causes and resulting symptoms. In Part II (Wakefield, 1997, Behaviour Research and Therapy, 35, 651-665, I examine Eysenck's (1986, Contemporary directions in psychopathology: Toward the DSM-IV) proposal for a dimensional diagnostic system to supplant DSM-IV.