The charts of 50 patients with the Dandy-Walker Syndrome were reviewed. Initial therapy was cystoperitoneal (CP) shunt in 21 (42%), ventriculoperitoneal (VP) shunt in 13 (26%), and both (CPVP) in 7 (14%), with the remaining 9 (18%) requiring no shunting procedure. Conversion from single to double shunt secondary to expansion of the unshunted compartment occurred in 9 (42%) of CP only and 4 (30%) of VP only patients. Final therapy was CP in 12 (24%), VP in 9 (18%), CPVP in 20 (40%) and none in 9 (18%). Patients who were unshunted remained shunt-free. When CPVP shunts were present, there was a statistically significant probability that shunt malfunction was due to the CP component only or both components, but not the VP component only. However, there was no statistical difference in malfunction or complication rates between patients with CP only or VP only shunts. Associated anomalies, intellectual outcome, and mortality were also analyzed. Seizures, hearing or visual problems, various CNS abnormalities, and systemic abnormalities were associated with poor intellectual development and could be used to predict intellectual outcome.