Studies using a variety of diagnostic rating scales have identified much higher levels of somatic symptoms in Hispanic psychiatric patients, particularly those with a diagnosis of schizophrenia or depression, than in their Anglo counterparts. However, this cultural difference cannot be generally demonstrated by using DSM-III somatoform disorders criteria. Many patients with "unfounded" somatic symptoms have another DSM-III primary diagnosis that is often assumed to cause or explain the somatic symptoms. In addition, the true prevalence of unfounded somatic symptoms is hidden because of the strict criteria required for a DSM-III primary diagnosis of somatization disorder. The author believes a more inclusive notion of somatization should be developed that would substantiate the presence of this trait among different populations and, if operationalized, would facilitate research on this important phenomenon. This is relevant not only because of the cross-cultural variability of somatization traits but also because they may become a salient feature of the primary syndrome and determine the use of clinical services. He proposes a new operational definition of somatization that encompasses the trait concept.