Thirty-eight insertions or revisions of ventriculopleural shunts were performed in 22 patients. The most common indication for pleural cerebrospinal fluid diversion was recent abdominal infection. Very minor pneumothoraces were seen frequently, but no patient became symptomatic or required treatment. Only three shunts in 2 patients were discontinued because of pleural effusions. The median survival of the entire series of ventriculopleural shunts was only 10 months. A historical control group of revised ventriculoperitoneal shunts matched for factors that have been shown to influence shunt survival was constructed from an institutional data base. The survival of simple ventriculopleural shunts in this series was not significantly different from that of simple ventriculoperitoneal shunts in patients of comparable age with a comparable recent shunt revision history (p = 0.68). Pleural cerebrospinal fluid diversion is a simple, safe, and reasonably effective option when clinical circumstances present relative or absolute contraindications to peritoneal diversion.