This article focuses on national data relevant to the following questions: (a) To what extent is mental hospitalization effective? (b) How many people are hospitalized for mental disorders yearly? (c) Has the rate of mental hospitalization been stable or rising in recent years? (d) What is the length of an inpatient episode and is it changing? (e) Has the rate of rehospitalization been rising? (f) What proportion of total • hospital days are for mental disorders? The article suggests that a set of beliefs or myths about mental hospitalization exists that is not supported by national data. The national data base on mental hospitalization is surprisingly sparse, but available data suggest the following: The rate of mental hospitalization is increasing, but the place in which it occurs has changed dramatically. The length of hospital stay has been stable in most sites over the last decade; a substantial decrease has occurred in two sites, but they handle only 20 % of the episodes. At least several treatment modes exist that appear to be more effective and less expensive than hospitalization. There is little evidence to support the notion of a "revolving door." Mental hospitalization is a substantial national issue: 25 % of all hospital days in the United States are for mental disorders. Issues related to the need for new public policy and the difficulties in changing curfenf policies are discussed.I have distinguished before between de jure and de facto national mental health policy (Kiesler, 1980). The de jure policy is that which we legislatively and collectively intend to carry out in the name of mental health. The de facto policy is that which occurs, regardless of public intent or agreement.In the case of mental health, the distinctions between de jure and de facto policies are very important. Our national de jure policy is the development of outpatient care and deinstitutionalization. The policy of developing outpatient care, at least, has been quite successfully implemented.