The treatment of optic neuritis (ON) with corticosteroids has been controversial for many years. None of the previous studies with the usual dosage of steroids or ACTH has proven that this therapy is beneficial. However, these studies did not apply sensitive parameters for assessing the course of ON. The aim of this study was to examine the effect of corticosteroids using subtle parameters. The study design emphasized: 1. strict criteria for patient selection; 2. frequent controls during the initial period of the disease, follow-up for 1 year; 3. double-blind conditions with randomized groups; 4. assessment of the course with sensitive parameters, including visual acuity, visual evoked potentials (VEP) and the Tübingen Flikker Test, also called the Aulhorn Flicker Test, which monitors the activity of the ON. These last two tests were used for the first time in a controlled treatment study in ON. Thirty-eight patients with acute ON were treated orally in a double-blind procedure with either methylprednisolone (initially 100 mg daily, reduction every 3 days) or with vitamin B1 (100 mg thiamin daily) in the control group. The results of the Flicker Test and VEP showed a trend toward faster recovery under prednisolone within the first few weeks after onset of the disease. After 1 year none of the parameters examined showed any difference between the two groups. Thus, although prednisolone treatment in the dosage mentioned above seems to speed up recovery in the initial phase in some patients, no long-term benefit could be demonstrated.