This is a case-control study of 160 persons with glioblastoma and 128 of their “best friends” as controls. Subjects came mainly from greater Boston, and data were gathered by questionnaire and telephone interview. Among those who had had a “severe” head injury at age 15 or later, the age-adjusted rate ratio (RR) of glioblastoma was 10.6, p = 0.004. There were six cases and no controls who had seizures for 15 or more years. The related RR is inestimable, but has a p value of 0.03. We could not evaluate whether the latter association implies a direct relationship between the causes of seizures and the causes of glioblastoma, or if it reflects the effect of another factor, such as medications to control the seizures.