Decompressive craniotomy (DC) in children is a life-saving procedure for the treatment of refractory intracranial hypertension related to traumatic, ischemic and infectious lesions. Different surgical procedures have been proposed including uni or bilateral hemicraniectomy, bi-frontal, bi-temporal, or bi-parietal craniotomies. DC can avoid the cascade of events related to tissue hypoxia, brain perfusion reduction, hypotension and the evolution of brain edema that can be responsible for brain herniation. The monitoring of intracranial pressure (ICP) is very important to take a decision as well as the value of Trans cranial Doppler (TCD). Repeated TCD in the intensive care unit give important information about the decrease of the cerebral perfusion pressure (CPP) and facilitates the decision making. The important question is about how long time we have to wait before to perform the DC. Three conditions can be distinguished: 1) ICP stable and TCD with good parameters: the decision can be postponed; 2) ICP>20 mmHg with good TCD and without clinical signs of deterioration: the decision can be postponed; 3) ICP>20 mmHg with altered CPP and degraded TCD value and clinical signs of brain herniation: the surgical procedure is indicated. The decision of a ventricular drainage can also be discussed but in cases of slit ventricles it is preferable to realize a DC to avoid the problems of multiple taps without finding the ventricular system. In some very specific situations, DC has to be contraindicated. The first one is a prolonged cardiopulmonary arrest with a no-flow longer than 15 minutes and irreversible lesions on the TCD or on the CT-scan. The second most common situation is a patient with GCS of 3 on admission associated with bilaterally fixed dilated pupils. In this case TCD is very useful to document the decrease or the absence of diastolic flux that indicates a very poor cerebral perfusion. In case of severe polytraumatism with multiorgan failure, especially in very severe hemorrhagic shock with incontrollable coagulopathy, the realization of DC is definitely hazardous with y a high risk of cardiac arrest during the surgical procedure. The decision to realize a hemicraniectomy or a bi-frontal craniotomy is related to the presence or not of associated traumatic lesions as hemorrhagic contusions or a sub-dural or extradural hematoma. In cases of diffuse cerebral edema the bi-frontal bone flap is indicated. In all cases a closure of the dura mater with a large dural patch has to be performed avoiding compression of the nervous system. Our results showed a mortality rate of 18%. Eighty percent of the survivors have a good quality of life but only 43% in a scholar age could attend a normal program. Patients treated with DC need a long follow-up and an important rehabilitation program to improve their quality of life. Our report shows that DC in children is effective to control the post-traumatic intracranial hypertension but a long follow-up is recommended to access the sequels and quality of life of these patients.