Analysis of two sets of steady-state data on bromide ion in the serum and cerebrospinal fluid of one human subject and of recently published data on the entry of bromide and sodium ions into human cerebrospinal fluid suggests that bromide is secreted out of the bulk of the cerebrospinal fluid, not excluded from it at a source, and that the secretory activity reaches a maximum at high concentrations of bromide ion. Submitted on August 25, 1958