An increasing number of severely ill patients require long-term care in non-hospital Residential Facilities (RFs). Despite the magnitude of this development, only a few detailed longitudinal studies on fairly larger samples of residents have provided important information about this population. The PERDOVE study is the first study carried out in Italy aimed at obtaining comprehensive data about the course and outcome of patients living in RFs, and test whether discharge to independent accommodations is a real option for many patients. The major aims are three: (1) to describe the sociodemographic, clinical, and treatment-related characteristics of patients treated in RFs during an index period in 2010; (2) to identify predictors, and characteristics associated to discharge at 1-year follow-up; and (3) to evaluate clinicians' predictions about the likelihood of home discharge for each individual patient. The primary endpoint was discharge within one year as a measure of good outcome. The PERDOVE study is an observational cohort study, involving 23 RFs all belonging to the St John of God Order, located in Northern Italy. The first step involves the evaluation of all RFs with a Structure Schedule, which assesses the following areas: logistics structure, general organization: internal rules, meals, clothing and personal hygiene; staff and users' characteristics and meetings, documentation system and evaluation activities; rehabilitative activities provided inside the RF itself. In the second step a research assistant completed a standardised assessment of each resident with the manager and staff; clinical diagnoses were made by treating clinicians. A number of standardised instruments were administered. At one year follow-up, changes in clinical and life conditions of the patients were evaluated and selected rating scales were administered again. For patients discharged to the community or to other RFs, a research assistant contacted the patient's psychiatrist and asked him/her to fill out the follow-up documentation. With regard to residents' discharge, investigators hypothesised that the likelihood of community discharge would be associated both with the level of social support outside the facilities, and with the degree of patients' psychological and social impairment.