Although epilepsy is commonly associated with shunt-treated hydrocephalus, its relation to the shunting procedure and the criteria identifying postoperative epilepsy remain controversial. Of 283 patients shunted at Würzburg University Hospital over a 24-year period (1970 to 1994), 182 were followed up for a minimum of 1 year after shunt insertion and entered the study. The data were analyzed retrospectively in 1995 and 1996. Epilepsy was analyzed in relation to the etiology of hydrocephalus, functional status, time and site of shunt insertion, onset of seizures and seizure type, EEG changes, sex, shunt systems, and shunt revisions. Of the 182 patients studied, 37 (20%) developed epilepsy. The incidence of epilepsy varied according to the etiology of hydrocephalus: posthemorrhagic (5%), postinfectious (4%), connatal/miscellaneous/unknown (3%), myelomeningocele (2%), tumor/arachnoidal cyst/aqueduct stenosis (0%). Early shunting and poor functional status was associated with a higher risk for epilepsy. Epilepsy was not influenced by sex, shunt systems, or number of shunt revisions. Twenty-two (12%) of 182 patients developed epilepsy (generalized N=13, focal N=9) after intracranial shunting. Focal EEG abnormalities (N=16) were located mainly at the anatomical site of the shunt (N=14), but only three patients (2%) presented with focal seizures contralateral and focal EEG abnormalities ipsilateral to the site of the shunt. The presence of epilepsy was determined by the etiology of hydrocephalus rather than by surgical intervention. The incidence of postoperative epilepsy (12%) was low. Onset of epilepsy, clinical presentation of seizures, and EEG changes did not appear to be valid criteria for identifying shunt-related epilepsy. Thus, epilepsy as a complication of intracranial shunting might be overestimated in the literature.