We implemented ventricular infusion studies on 33 patients suspected of idiopathic normal pressure hydrocephalus (iNPH), benign intracranial hypertension (BIH) or occlusive hydrocephalus (HOC) in order to confirm shunt indications. The initial scope was to study O(2) supply during infusion tests to exclude further violation of already vulnerable brains during ICP elevation. Intraventricular infusion was performed via ventricle catheters with the ICP tip sensor, while brain tissue oxygenation was measured with intraparenchymal Raumedic PTO probes. In 15 out of 23 (65%; 8 NPH, 2BIH, 5 HOC), pO(2) increased constantly (average 140%), while brain temperature decreased (range: 0.2-4.5°C) during the infusion studies. In another six patients, O(2) values remained largely stable during the infusion studies (4NPH, 1BIH, 1HOC). Cerebral deoxygenation during infusion tests occurred only in two patients (1NPH, 1HOC).Overall cerebral oxygenation and temperature inversely correlated well with some temporary delay regarding oxygenation state as a consequence of cerebral temperature. Probably, this effect is a consequence of reduced cerebral metabolism caused by local cooling. We hypothesise that such cooling is mediated via the large basal arteries and suggest that such a pathophysiology, ICP-controlled local cooling, might offer a new option for brain protection (e.g. in an ICP crisis).