OBJECT: The authors analyzed the role of endoscopic third ventriculostomy (ETV) as a primary treatment for hydrocephalus and also as an alternative to shunt revision for malfunctioning and infected ventriculoperitoneal (VP) shunts. METHODS: A retrospective analysis of clinical notes, operation records, and magnetic resonance imaging procedures before and after ventriculostomy was performed to determine the success or failure of ETVs in 170 patients who underwent a primary ETV and in 63 patients who underwent an ETV for shunt malfunction (secondary ETV). The patients' data were derived from an endoscopy database inaugurated in 1998. Of the 63 patients with shunt malfunctions, 49 patients (78%) had mechanical malfunction only and 14 patients (22%) had both infection and malfunction. Seventy-four percent (126 of 170) of patients in the primary ETV group and 70% (44 of 63) of patients in the secondary ETV had a successful outcome at the time of analysis. The success rate for ETVs in cases involving a mechanical shunt malfunction alone was 67% (33 of 49) compared with 79% (11 of 14) in those cases involving an infected shunt. The origin of the hydrocephalus in the primary and shunt malfunction groups was evaluated as a factor contributing to the success of the ETV. In the primary group, patients with a history of intraventricular hemorrhage (IVH) and meningitis as a cause for hydrocephalus had a poor rate of success after the ETV--27% (four of 15) and 0% (none of two), respectively. This pattern was not seen in the series involving shunt malfunction after the ETV, with 71% (five of seven) and 75% (three of four) of cases having a hydrocephalus origin of IVH and meningitis, respectively, in which a successful outcome was attained. A two-way mixed-model analysis of variance yielded a significant effect for origin (p = 0.011), a significant interaction between group and origin (p = 0.028), and a marginally nonsignificant effect of group (p = 0.0686). More than 95% of failures were evident within 1 month of the ETV in both groups. Complications were minimal in both groups, and there were no procedure-related deaths. CONCLUSIONS: An ETV is a safe procedure with few complications and a high success rate in both primary and secondary groups. An ETV to address shunt malfunction, unlike a primary ETV, is not particularly origin specific. A bonus is its success in dealing with infected shunts. Most failures will be evident early, but long-term follow up is vital.