A comparison of intraventricular pressure (IVP) and extradural pressure (EDP) was carried out in 17 patients being investigated for normal-pressure hydrocephalus, and in six patients with acutely raised intracranial pressure following events such as head injury or intracerebral hematoma. Extradural pressure was measured using the CardioSearch monitor. There was a reasonably good correlation between EDP and IVP in the chronic stable group with pressures up to 25 mm Hg. In the acute group there was no predictable relationship between EDP and IVP, and during a 24-hour period the pressures could vary by as much as 30 mm Hg. Subdural pressure, measured with the same instruments, was compared to IVP in both acute and stable situations in eight other patients: there was a close and constant correlation between pressures in these two spaces. The authors conclude that misleading information may be obtained from EDP monitoring, and erroneous management decisions may result from dependence on such a technique. Possible explanations for this are discussed.