Neurotoxicity, especially associated with therapy for acute lymphatic leukemia, has been attributed by many to the use of methotrexate (MTX). After radiotherapy this has been reported even more frequently but no explanation has been suggested apart from "a complex interaction". The hypothesis presented here is as follows: (1) Inadequate folinic acid rescue alone accounts for MTX-induced neurotoxicity. (2) Adequate folinic acid after MTX can prevent neurotoxicity. (3) Higher doses of MTX require a disproportionately higher dose of folinic acid than MTX to prevent toxicity. Doubling the dose of MTX has required tripling and quadrupling the folinic acid dose to prevent neurotoxicity. Assuming that central nervous system radiotherapy increases the cerebrospinal fluid levels of MTX and folinic acid proportionally, the folinic acid level may now not be enough to prevent neurotoxicity. This neurotoxicity occurs when MTX is given after (but not before) radiotherapy, and can be prevented by appropriate doses of folinic acid. (4) There is no evidence that within the dose range currently being used, the administration of larger quantities of folinic acid to prevent neurotoxicity compromises prognosis. This hypothesis is supported by a large amount of published data. Critical reanalysis of studies that ostensibly contradict parts of the hypothesis showed that they, in fact, support it.