OBJECTIVE: To investigate the recently defined Positive and Negative Syndrome Scale for Schizophrenia remission criteria in a naturalistic setting of psychotic patients; to identify causal factors that change remission status; and to validate the criteria against global indices of illness, cognitive functions, and social outcome. METHODS: This was a longitudinal naturalistic study of 162 patients, diagnosed with schizophrenia or schizophrenia-related psychotic disorders (mean illness duration, 11 years) and treated with risperidone at study entry. Symptoms, drug treatment, cognitive function, and social outcome were measured at baseline and annually for 5 years. Remission was constructed retrospectively with only indirect data on stability over time. RESULTS: At study entry, 40% of the patients with schizophrenia were in symptomatic remission, stabilizing between 55% and 60% after a few years. The need for hospitalization became less frequent over time; initially 31%, dropping to 7% by years 4 and 5. Many patients went in and out of remission. Remission was strongly associated with global indices of illness, with intact insight and with social outcome (except work/studies) but not with cognition or medication. CONCLUSIONS: In spite of certain weaknesses of the study, we may conclude that current definition of remission is primarily a symptomatic measure, covering a subset of symptoms, some of which are not schizophrenia-specific. Although the definition may be clinically relevant, we must be aware of the risk that the connotation of the word could induce too much focus on symptom control.