OBJECTIVE: A single-center, retrospective study was performed to evaluate the effect of audit on the patient selection for endoscopic third ventriculostomy (ETV). MATERIALS AND METHODS: Between 01.01.99 and 07.31.01, 134 patients underwent ETV (group 1). During this period, there was no consensus within the neurosurgical community as to patient selection criteria for ETV. A review of our clinical practice in August 2001 demonstrated significantly lower ETV success rates for patients <6 months of age, patients with communicating hydrocephalus (HC) and for patients with prior shunt surgery. Thus, stricter patient selection criteria were established. Between 08.01.01 and 12.31.02, 54 patients were operated (group 2). The two groups were compared with respect to age, type of HC, previous shunt surgeries and ETV success rates. The primary outcome event was ETV malfunction, defined as symptoms and/or signs of increased intracranial pressure leading to repeat ETV or shunt implantation. Follow-up was done through outpatient clinics and telephone interviews. Average follow-up time was 12 months (range 0-44 months). No patient was lost to follow-up. RESULTS: The overall 1-year ETV success rate in group 2 (65%) was significantly higher than in group 1 (53%) (p < 0.04). Group 2 had a significantly higher proportion of patients >6 months of age (p = 0.013) and with obstructive HC (p = 0.001). CONCLUSION: Patient selection criteria critically affect the overall ETV success rate. An audit of our results led to a significant change in clinical practice, thereby improving the ETV success rates and patient care.