Very preterm infants are prone to numerous medical complications with lifelong impact. Amongst the most serious are severe intraventricular hemorrhage (sIVH) and the subsequent progression to posthemorrhagic hydrocephalus (PHH). Currently, the only treatment for PHH is surgery, most commonly with shunts that are prone to malfunction across the lifespan. Children with sIVH and PHH are also at high risk for intellectual disability, behavioral problems, neurosensory impairment, cerebral palsy and epilepsy. Emerging evidence suggests that the cellular and molecular events regulating cerebrospinal fluid (CSF) dynamics, including CSF secretion, propulsion and reabsorption, develop during the third trimester and the first few months postnatally. Maturation of these highly subspecialized and metabolically active cellular processes spatially and temporally overlaps with preterm birth and IVH, and are thus vulnerable to injury over an extended period. Most importantly, these processes are responsive to neurorestorative interventions with re-purposed medications, raising the possibility of using medical treatment after sIVH to prevent progression to PHH and the need for shunts. Preclinical data show that melatonin (MLT) and erythropoietin (EPO), when administered in a sustained dosing regimen, can prevent the hallmarks of progression from early postnatal sIVH to subsequent PHH, including macrocephaly and ventriculomegaly. Combination therapy is required as neither agent alone prevents PHH. In human preterm infants, MLT and EPO have been safely used as monotherapy in clinical trials with similar dosing regimens. We propose a Phase I, single institution, randomized, double-blind trial for very preterm infants with sIVH to define a safe combination dose of MLT and EPO. With IRB, IND and primary neonatologist approval, and informed consent, a maximum of 60 very preterm neonates with sIVH will be enrolled, treated through 33w6/7d, and followed to 37w6/7d. Neonates will be randomized 3:1 between MLT+EPO and placebo, with all receiving standard of care. Concurrent controls are needed due to fluctuations in preterm birth co-morbidities and mortality. Masking is essential to reduce attribution bias. The primary endpoint is a composite serious adverse event (SAE)/dose limiting toxicity (DLT) including death, potential MLT-realted DLT: severe liver function abnormalities compared to age-matched peers with sIVH, and known EPO-related SAE: thrombosis, polycythemia, and hypertension. No MLT-related SAE have emerged in clinical trials thus far. We hypothesize that the MLT+EPO SAE/DLT rate will not be higher than the placebo rate. Secondary outcomes will be rate of co-morbidities of preterm birth. Exploratory data, collected to guide design of future clinical trials for efficacy, will include serial neuro-imaging metrics acquired from clinical images, serial neonatal neurodevelopmental examinations, serum and urine MLT and EPO levels, and liquid biomarkers. Successful impl...