There is evidence that some event in childhood may determine risk of multiple sclerosis: Elevated titers to measles and other childhood infections suggest a childhood infection. Therefore, childhood infections reported by 30 patients with multiple sclerosis and matched controls were compared. Patients reported a childhood infection between 5 and 9 years (not simply exposure to an infection) more often than controls. The mean age of measles peaked somewhat later (age 7) in patients than in controls (age 4); this differnce approached statistical significance (p less than 0.1). Evidence that host response to measles is age-dependent was reviewed. It was proposed that age of measles (rather than the fact of injection) may influence the risk of developing multiple sclerosis.