A mortality-based case-control study of selected risk factors for childhood brain tumors was undertaken. Ohio-born children who died from brain cancer during the 1959-1978 vicennium were compared to control children (of the same age, race, and sex) by using information obtained from the subjects' birth certificates. Differences between the case and the control children with respect to paternal occupation, the focus of the study, were examined. Controlling for the potentially confounding effects of several nonoccupational factors, case fathers were found more likely than control fathers to have been employed (at the time of birth of their children) in agriculture, in metal-related jobs, in structural work jobs in the construction industry, and in electrical assembling, installing, and repairing occupations in the machinery industry. Although the results must be interpreted with caution, the findings lend support to the hypothesis that parental occupation is a potential risk factor for childhood brain tumors.