Data on all obstetric patients delivering at the Brigham and Women's Hospital during the years 1982 through 1987 were collected. The anesthetic techniques used, the type and amount of anesthetic agents administered, and the postpartum relapse rate of multiple sclerosis patients were compared. Women who received epidural anesthesia for vaginal delivery did not have a significantly higher incidence of relapse than those who received local infiltration. However, all of the women who experienced postpartum relapses had received concentrations of bupivacaine greater than 0.25%. This finding may suggest that a higher concentration of drug over a longer period of time may adversely influence the relapse rate.