There is no firm evidence regarding cerebrospinal fluid (CSF) shunt reimplantation after infection in the pediatric population. The purpose of this study was to compare different criteria and analyze new shunt failure. A cross-sectional retrospective multicenter study was performed over 6 years to study patients and each infected shunt at diagnosis, reimplantation, and after reimplantation. The patients were divided into 2 groups: group 1 (G1), reimplantation after negative serial CSF cultures during antibiotic treatment; group 2 (G2), reimplantation after negative serial pancultures after completion of antibiotics. The differences were measured with Mann-Whitney and Χ2 tests; multivariate analysis and associations were calculated using odds ratios (ORs) based on logistic regression. There were 137 shunt infection events in 110 patients: 28 events in G1 and 109 in G2. Significant differences were observed in the diagnosis and reimplantation. Reimplantation dysfunction in G1 was 16 (55.17%) versus 30 (27.78%) in G2 (P= 0.006). The risk of shunt malfunction after reimplantation increased for G1 reimplantation criteria (P= 0.018; OR, 3.34; confidence interval [CI], 1.23-9.05): pleocytosis at diagnosis >17 cells (P= 0.036; OR, 2.41; CI, 1.06-5.47), CSF proteins at diagnosis >182 mg/dL (P= 0.049; OR, 2.21; CI, 1.00-4.89). G2 reimplantation criteria were related to improved pleocytosis, CSF proteins, and blood neutrophils compared with G1. Mechanical and infectious dysfunction of the new shunt was 3 times more prevalent in G1 than in G2, considering the differences between the groups at diagnosis. Increased parameters of infection at diagnosis were associated with future malfunction more than parameters before reimplantation in both groups.