The effects of therapeutic prednisolone treatment on experimental allergic neuritis (EAN) in Lewis rats were evaluated in a controlled clinical and electrophysiological study. Since steroid therapy has been suspected to cause relapsing or chronic disease, monitoring was extended over 200 days. Short-term steroid treatment (5 days of 15 mg/kg prednisolone, n = 8) with sudden steroid withdrawal was compared with long-term application (30 days, beginning at 7.5 mg/kg) in descending dosage (n = 8). The experiment included saline-injected controls (n = 8) and controls for stress possibly exerted by the handling of the animals. Treatment was begun at the onset of clinical signs. The clinical and electrophysiological data indicated that deterioration, recovery and mild (insignificant) relapse (after day 30 and day 108) occurred in all groups at the same time. Both steroid application schemes significantly (p less than 0.03) attenuated the severity and shortened the duration of EAN. Relapse was not aggravated after steroid treatment. The clinical course and electrophysiological findings were unaltered by the experimental procedures and by mild experimental stress.