Ventricular enlargement in schizophrenia as measured by ventricular-brain ratio (VBR) has been described by three different research groups. Because of the possibility that the samples might not be representative of the broad population of schizophrenic patients, questions have been raised about generalizability. The authors report on ventricular enlargement in a rigorously defined but representative sample of schizophrenic patients intermittently hospitalized but living in the community. They provide additional evidence that ventricular enlargement does occur in schizophrenic patients, although not at rates as high as those previously reported. Issues related to the definition of ventricular enlargement, as well as the effects of sampling and choice of a control group, are discussed.