To study the incidence of circulating anti-CNS antibodies in childhood neurologic diseases, a population study was undertaken. Serum samples were obtained from a total of 348 children and stored at -80 degrees C until being studied. The samples were collected when routine blood tests were being performed. In all cases informed consent was obtained. This study was approved by hospital ethics review committees. One hundred and ninety-nine of the samples were from children with no known neurologic illnesses and served as the control group. One hundred and twenty-one of the samples were from children with epilepsy and the remaining 28 from a number of different neurologic conditions. The serum samples were screened against normal, adult, autopsy-derived cerebellar and frontal cortex tissue sections and Western blots. Serum immunoreactivity was revealed using HRP-conjugated anti-human IgG. Significant findings included: (1) patients with epilepsy had an increased incidence of anti-CNS reactivity as revealed on frontal cortex immunoblots (p less than 0.05) but not on cerebellar immunoblots; (2) there was an increase in the incidence of immunoblot reactivity with age in the controls and the neurology cases; (3) there was an increased incidence of immunoblot reactivity in those cases with a presumed inflammatory central or peripheral neurologic disease; (4) in six additional cases with opsoclonus-myoclonus there was cerebellar-specific immunoreactivity with identified antigenic molecular weights of 27 and 35, and 62 kDaltons; (5) in 31 additional cases of systemic lupus erythematosus there was significant immunoblot reactivity (p less than 0.001) when compared to a subset of age-matched controls. There was no difference in immunoreactivity between males and females. There was no significant increase in immunoreactivity in those children with cognitive disturbances including developmental delay and mental retardation.