Some studies indicate that infants, especially those less than 1 month of age have a higher incidence of ventriculoperitoneal shunt infections. To look at age as well as other variables that might relate to the rate of shunt infection, we reviewed the records of all patients undergoing a ventriculoperitoneal shunt insertion or revision at our institution from January 1, 1985, to December 31, 1994. There were a total of 2,325 ventriculoperitoneal shunting procedures performed on 1,193 patients with a male:female ratio of 678:515. The overall infection rate was 3.2% (74 infections). Analyzed by age, the infection rates were as follows: <1 month 9/223 (4.0%), 1-6 months 16/449 (3.6%), 6-12 months 13/297 (4.4%), 12-18 months 3/122 (2.5%), 18-24 months 7/116 (6.0%) and 24+ months 26/1, 118 (2.3%). There was no statistically significant difference between age groups (p > 0.05). Upon selectively examining premature neonates who developed hydrocephalus secondary to intraventricular hemorrhage from the figures given above, one finds that 2/44 (4.5%) of neonates became infected, which was also not significant. The infection rate was the same irrespective of whether the procedure was to insert or revise the shunt, or whether another operative procedure was done under the same anesthesia. The etiology of the hydrocephalus was not a factor, nor was the presence of an open neural tube defect. The presence of fluid accumulation along the shunt tract or at another neurological operative site was associated with a significant increase in incidence of infection 15/168 (8.9%) when compared to those with no fluid accumulation (p < 0.001). The type of infecting organism was divided roughly in thirds, with relatively equal representation from Staphylococcus epidermidis/coagulase negative and Staphylococcus aureus. The remaining third was comprised of a wide variety of organisms.