Five patients with hydrocephalus were studied with carbon-11-2-deoxyglucose or 2-deoxy-2-(18F) fluoro-D-glucose and positron emission tomography both prior to and following ventricular shunting. Four subjects had communicating hydrocephalus; the fifth had aqueductal stenosis, two patients had hydrocephalus for three months or less. The three chronic patients were felt to have hydrocephalus for three years or more. After shunting ventricular size decreased in all patients, and all patients showed clinical improvement. The glucose cerebral metabolic rates increased after shunt in the two subjects with recent onset hydrocephalus but paradoxially decreased in the three chronic patients despite clinical improvement. These findings suggest that the cerebrum was metabolically hyperactive prior to shunt due to an unknown mechanism and presumably in response to the presence of hydrocephalus. A dissociation may also exist in the post-shunt period between cerebral metabolism and cerebral blood flow.