One hundred eighteen epileptic children, aged 6-15 years, underwent detailed neuropsychological testing including the Wechsler Intelligence Scale for Children-Revised and the age-appropriate Halstead-Reitan battery. Eight had classical absence seizures only, eight had classic absence seizures and generalized tonic-clonic seizures, 30 had generalized tonic-clonic seizures only, 31 had partial seizures only, 20 had partial seizures and generalized seizures, 15 had atypical absence seizures, and five had minor motor seizures. A control group of 100 children without seizures, matched to the general population for intelligence and matched to the seizure cases for age, underwent identical testing. The Wechsler full-scale intelligence quotient (FSIQ) of cases was significantly (p = 0.01) lower than that of controls and was related to seizure type. Children with minor motor or atypical absence seizures had the lowest average FSIQ (70 and 74, respectively). All seizure types were associated with below-control intelligence except classic absence only. Intelligence was also correlated with degree of seizure control. A highly significant inverse correlation between years with seizures and intelligence was found (p less than 0.0001). A rating of neuropsychological impairment, derived from all measures of brain function, was assigned to each child. Epileptic children had significantly more impairment than controls (p less than 0.0001). Children with seizures had been placed in special education or had repeated a grade in school almost twice as frequently as controls (p less than 0.001). Even though often placed in a class of younger children, their academic achievement was behind grade placement more often than in controls.(ABSTRACT TRUNCATED AT 250 WORDS)