Posthaemorrhagic hydrocephalus (PHH) is a major problem for premature infants, generally requiring lifelong care. It results from small blood clots inducing scarring within CSF channels impeding CSF circulation. Transforming growth factor – beta is released into CSF and cytokines stimulate deposition of extracellular matrix proteins which potentially obstruct CSF pathways. Prolonged raised pressures and free radical damage incur poor neurodevelopmental outcomes. The most common treatment involves permanent ventricular shunting with all its risks and consequences.This is a review of the current evidence for the treatment and prevention of PHH and shunt dependency. The Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library) and PubMed (from 1966 to August 2008) were searched. Trials using random or quasi-random patient allocation for any intervention were considered in infants less than 12 months old with PHH. Thirteen trials were identified although speculative interventions were also evaluated.The literature confirms that lumbar punctures, diuretic drugs and intraventricular fibrinolytic therapy can have significant adverse effects and fail to prevent shunt dependence, death or disability. There is no evidence that postnatal phenobarbital administration prevents intraventricular haemorrhage (IVH). Subcutaneous reservoirs and external drains have not been tested in randomized controlled trials, but can be useful as a temporising measure. Drainage, irrigation and fibrinolytic therapy as a way of removing blood to inhibit progressive deposition of matrix proteins, permanent hydrocephalus and shunt dependency, are invasive and experimental. Studies of ventriculo-subgaleal shunts show potential as a temporary method of CSF diversion, but have high infection rates.At present no clinical intervention has been shown to reduce shunt surgery in these infants. A ventricular shunt is not advisable in the early phase after PHH. Evidence exists that pre-delivery corticosteroid therapy reduces mortality and IVH and there may be trends towards reduced disability in the short term. There is also evidence that postnatal indomethacin reduces IVH but with no effect on mortality or disability. Overall, there is still no definitive algorithm for the treatment of PHH or prevention of shunt dependence. New therapeutic approaches in neonatal care, including those aimed at pre-empting PHH, offer the best hope of improving neurodevelopmental outcomes.