A battery of neuropsychologic tests was administered "blindly" to 18 children with acute lymphocytic leukemia (ALL) who had been randomly assigned to treatment regimens with or without cranial radiation. These children were all in complete continuous remission for more than 3 1/2 years and were no longer receiving therapy. The results indicated no substantial differences between groups as a function of radiation therapy. However, decreased neuropsychologic performance was found when the entire sample was compared with population norms. These data do not support the hypothesis that cranial radiation therapy is responsible for the neuropsychologic sequelae seen in these survivors of ALL. Post hoc multiple regression analysis indicated that parental education levels accounted for more of the neuropsychologic variability seen in these children than other factors such as age at diagnosis, type of therapy, or sex of child.