OBJECTIVE: We report on 32 neonates treated with ventriculosubgaleal (VSG) shunts to determine VSG shunt survival and associated complications. METHODS: Between 1993 and 1997,37 VSG shunts were placed in 32 neonates when the cerebrospinal fluid (CSF) or the abdomen was considered unsuitable for ventriculoperitoneal shunt placement. In each child, a ventricular catheter was attached to 3 cm of a closed-end peritoneal tube via a right-angle connector, which drained into a surgically created subgaleal pocket. RESULTS: The causes of hydrocephalus were as folows: intraventricular hemorrhage (IVH) in 20 neonates, meningitis/ventriculitis in 6, IVH and infection in 2, and other causes in 4. The mean postconception age at the time of VSG shunt insertion was 37.2 weeks (33.1 wk in the IVH group), and the mean weight was 2227 g (1724 g in the IVH group). The average preoperative head circumference was 33.6 cm. The average survival of these 37 VSG shunts (five children had two VSG shunts) was 35.1 days. The complications were as follows: one CSF leakage occurred when sutures were removed; one catheter fell into the ventricle and required removal, and one child died immediately after VSG shunt revision. There were no VSG shunt infections. All surviving children followed for a minimum of 4 months after insertion of a VSG shunt (n = 24) have required a ventriculoperitoneal shunt. Mean follow-up from the time of first VSG shunt was 21.6 months. Four children died as result of causes unrelated to the VSG shunt. CONCLUSION: VSG shunts offer a simple, effective, and relatively safe means of temporizing hydrocephalus, and they avoid the need for external drainage or frequent CSF aspiration in these medically unstable infants until the CSF characteristics and abdomen are acceptable for ventriculoperitoneal shunting.