The term "prodromal symptoms" has traditionally referred to prepsychotic changes in thought, affect, and cognition that precede the initial onset of schizophrenia. Recently, however, the term has been extended into a clinical action context to refer to the early warning signs (EWS) of impending relapse in patients already diagnosed as having schizophrenia. However, recent reports reviewed by Norman and Malla (1995, this issue) use a narrow definition of prodromal symptoms and question their use in the clinical action context. We argue that the dual use of the term "prodromal symptoms" has led to conceptual confusion and to the impression that EWS cannot be used effectively for clinical action. The ability to base clinical action on EWS is central to schizophrenia therapeutics and is the cornerstone of pharmacological strategies based on early intervention. Our review of the evidence suggests that the effective clinical use of EWS depends on (1) the inclusion of both psychotic and nonpsychotic symptoms as EWS; (2) the use of clinician judgment in combination with predefined symptom changes to define the occurrence of EWS; (3) frequent clinical visits; and (4) the use of family or caregiver informants. We therefore suggest that, in the clinical action context, the terminology "early warning signs of impending relapse" should be used instead.