OBJECTIVE Posthemorrhagic hydrocephalus (PHH) is associated with significant morbidity, smaller hippocampal volumes, and impaired neurodevelopment in preterm infants. The timing of temporary CSF (tCSF) diversion has been studied; however, the optimal time for permanent CSF (pCSF) diversion is unknown. The objective of this study was to determine whether cumulative ventricle size or timing of pCSF diversion is associated with neurodevelopmental outcome and hippocampal size in preterm infants with PHH. METHODS Twenty-five very preterm neonates (born at ≤ 32 weeks’ gestational age) with high-grade intraventricular hemorrhage (IVH), subsequent PHH, and pCSF diversion with a ventriculoperitoneal shunt (n = 20) or endoscopic third ventriculostomy (n = 5) were followed until 2 years of age. Infants underwent serial cranial ultrasounds from birth until 1 year after pCSF diversion, brain MRI at term-equivalent age, and assessment based on the Bayley Scales of Infant and Toddler Development, Third Edition, at 2 years of age. Frontooccipital horn ratio (FOHR) measurements were derived from cranial ultrasounds and term-equivalent brain MRI. Hippocampal volumes were segmented and calculated from term-equivalent brain MRI. Cumulative ventricle size until the time of pCSF diversion was estimated using FOHR measurements from each cranial ultrasound performed prior to permanent intervention. RESULTS The average gestational ages at tCSF and pCSF diversion were 28.9 and 39.0 weeks, respectively. An earlier chronological age at the time of pCSF diversion was associated with larger right hippocampal volumes on term-equivalent MRI (Pearson’s r = −0.403, p = 0.046) and improved cognitive (r = −0.554, p = 0.047), motor (r = −0.487, p = 0.048), and language (r = −0.414, p = 0.021) outcomes at 2 years of age. Additionally, a smaller cumulative ventricle size from birth to pCSF diversion was associated with larger right hippocampal volumes (r = −0.483, p = 0.014) and improved cognitive (r = −0.711, p = 0.001), motor (r = −0.675, p = 0.003), and language (r = −0.618, p = 0.011) outcomes. There was no relationship between time to tCSF diversion or cumulative ventricle size prior to tCSF diversion and neurodevelopmental outcome or hippocampal size. Finally, a smaller cumulative ventricular size prior to either tCSF diversion or pCSF diversion was associated with a smaller ventricular size 1 year after pCSF diversion (r = 0.422, p = 0.040, R2 = 0.178 and r = 0.519, p = 0.009, R2 = 0.269, respectively). CONCLUSIONS In infants with PHH, a smaller cumulative ventricular size and shorter time to pCSF diversion were associated with larger right hippocampal volumes, improved neurocognitive outcomes, and reduced long-term ventriculomegaly. Future prospective randomized studies are needed to confirm these findings.