We have attempted to survey the current practice of inpatient treatment for the new cohort of chronic patients and to view it through the lenses of new glasses. Most controlled outcome studies have shown that outpatient treatment, which includes partial hospitals and clinics, is equal in efficacy to inpatient treatment, and that outcomes for shorter treatment are equal to those for longer treatment except for the two subgroups noted.Based on the data, we have descnibed enabling factors, indications, and contraindications for hospitalization. Our belief is that some chronic patients require periodic rehospitalization, but only some, and in these cases the focus should be on highly specific, problem-oriented treatment, most of which can be accomplished in about 30 days. In part, this guideline challenges the traditional assumption that most acutely disorganized patients require inpatient care to provide central structure and treatment until they can funciion without such support. Even with these guidelines the decision of whether to hospitalize or not hospitalize the chronic patient will remain problematic.The most cost-effective and clinically sound approach is to view the hospital as one part of an array of comprehensive, continuous, and coordinated mental health services in which specific target problems can be diagnosed, managed, and treated in order to maintain the patient in the community. Adopting this view requires changes in clinical practices.