We compared prednisolone (PRED) and azathioprine (AZA) versus prednisolone alone in the treatment of MG. Prednisolone alone or combined with azathioprine is widely used in the treatment of MG, but no randomized placebo-controlled comparative trial data are available. The prednisolone dose and clinical outcome were compared in a multicenter randomized double-blind study of 34 MG patients who were followed up for 3 years. One group (PRED + AZA) received prednisolone (on alternate days) plus azathioprine (2.5 mg/kg); the other group received prednisolone on alternate days plus placebo (PRED + PLAC). Initial high-dose prednisolone (1.5 mg/kg on alternate days) was tapered at remission to the minimal dose required to maintain remission. The prednisolone dose did not differ significantly between the two groups at 1 year (median values: PRED + AZA, 37.5 mg on alternate days; PRED + PLAC, 45 mg on alternate days) but was reduced at 2 and 3 years in the PRED + AZA group (median value at 3 years: PRED + AZA, 0 mg on alternate days; PRED + PLAC, 40 mg on alternate days; p=0.02). Relapses and failures to remit over the 3 years were more frequent in the PRED + PLAC group. There was a sharp rise in the anti-acetylcholine receptor (AChR) titers in the PRED + PLAC group at 2 years. Incidence of side effects was slightly less in the PRED + AZA group. Azathioprine as an adjunct to alternate day prednisolone in the treatment of antibody-positive generalized MG reduces the maintenance dose of prednisolone and is associated with fewer treatment failures, longer remissions, and fewer side effects.