41 patients with pemphigoid entered a multicentre randomised study of the efficacy of plasma exchange. All patients received 0.3 mg/kg daily oral prednisolone, increased weekly if the disease remained active. 24 patients received plasma exchanges in addition (eight large-volume exchanges over 4 weeks), and 17 had prednisolone only. 4 patients, 2 in each group, were withdrawn from the study. The initial dose of prednisolone was effective in 13 of 22 patients receiving plasma exchange but in none of 15 patients receiving prednisolone only. Control of the disease was obtained with a mean daily prednisolone dose of 0.52 +/- 0.28 mg/kg in the plasma exchange group v 0.97 +/- 0.33 mg/kg in the other group and a mean cumulative dose of 1240 +/- 728 mg v 2770 +/- 1600 mg. This finding suggests that plasma exchange allows a substantial saving of corticosteroids in the management of pemphigoid. This sparing effect was observed whether or not serum anti-basement membrane antibodies had been detected before treatment.