Endoscopic third ventriculostomy (ETV) has gained traction as a method for treating post-hemorrhagic hydrocephalus of prematurity (PHHP) in an effort to obviate lifelong shunt dependence in neonates. However, data remains limited regarding inpatient failures. A retrospective analysis of the NIS between 1998 and 2014 was performed. Discharges with age < 1year and ICD-9-CM codes indicating intraventricular hemorrhage of prematurity (772.1x) and ETV/shunt (02.22 and 02.3x) were included. Patients with ICD-9-CM codes for ventricular drain/reservoir (02.21) were excluded to prevent confounding. Time trend series plots were created. Yearly trends were quantified using logarithmic regression analysis. Kaplan-Meier curves were utilized to analyze time to treatment failure. Time to failure for each treatment was compared using log-rank. A total of 11,017 discharges were identified. ETV was more likely to be utilized at < 29weeks gestational age (p = 0.0039) and birth weight < 1000g (p = 0.0039). Shunts were less likely to fail in older and heavier newborns (OR 0.836 p = 0.00456, OR 0.828 p = 0.0001, respectively). Those initially shunted had lower failure rates compared with ETV (OR 0.44, p < 0.0001) but time to failure was longer with ETV (p = 0.04562). 79.5% of ETVs that failed were shunted after the first failure. Shunts were much less likely to undergo ETV if they failed (OR 0.21, p < 0.0001). Higher grade IVH was predictive of shunt failure but not ETV (OR 2.36, p = 0.0129). Although ETV can be effective in PHHP, it has a much higher initial failure rate than shunting and should thus be chosen based on a multifactorial approach.