OBJECTIVE The objective of this investigation was to review the clinical characteristics of recurrent CSF shunt infections in a large pediatric neurosurgical practice and to assess the safety and efficacy of reusing original ventricular entry sites for external ventricular drainage during treatment of infections and for subsequent reinsertion of shunts. METHODS Prospectively accrued clinical data on all patients treated at Children's Hospital Colorado for CSF shunt infections within a 10.5-year span were retrospectively investigated. RESULTS One hundred twenty-one consecutive cases of CSF shunt infection met inclusion criteria. Recurrent shunt infection attributable to the management of these infections occurred in 14 cases (11.6%). Three recurrent infections were with their original organisms, 7 were organisms different from the original organisms, and 4 were indeterminate. CONCLUSIONS Half or more of recurrent shunt infections were with organisms different from the original organism, and hence were new-type infections introduced during the management of the original infections. Incomplete eradication of original pathogens accounted for 3 (21.4%) of the 14 recurrent infections. Reusing recently infected or contaminated ventricular entry sites, both for CSF drainage during treatment and for implantation of new shunts, was as safe, with regard to risk of recurrent infection, as switching to new entry sites. Prior evidence of shunt infection is not, alone, a sufficient reason to change to a previously well-functioning site, and reuse of contaminated ventricular entry sites avoids all risks associated with making new ventricular entries.