SummaryBackground High doses of intravenous methylprednisolone are recommended to treat relapses in patients with multiple sclerosis, but can be inconvenient and expensive. We aimed to assess whether oral administration of high-dose methylprednisolone was non-inferior to intravenous administration. Methods We did this multicentre, double-blind, randomised, controlled, non-inferiority trial at 13 centres for multiple sclerosis in France. We enrolled patients aged 18–55 years with relapsing-remitting multiple sclerosis who reported a relapse within the previous 15 days that caused an increase of at least one point in one or more scores on the Kurtzke Functional System Scale. With use of a computer-generated randomisation list and in blocks of four, we randomly assigned (1:1) patients to either oral or intravenous methylprednisolone, 1000 mg, once a day for 3 days. Patients, treating physicians and nurses, and data and outcome assessors were all masked to treatment allocation, which was achieved with the use of saline solution and placebo capsules. The primary endpoint was the proportion of patients who had improved by day 28 (decrease of at least one point in most affected score on Kurtzke Functional System Scale), without need for retreatment with corticosteroids, in the per-protocol population. The trial was powered to assess non-inferiority of oral compared with intravenous methylprednisolone with a predetermined non-inferiority margin of 15%. This trial is registered with ClinicalTrials.gov, number NCT00984984. Findings Between Jan 29, 2008, and June 14, 2013, we screened 200 patients and enrolled 199. We randomly assigned 100 patients to oral methylprednisolone and 99 patients to intravenous methylprednisolone with a mean time from relapse onset to treatment of 7·0 days (SD 3·6) and 7·4 days (3·9), respectively. In the per-protocol population, 66 (81%) of 82 patients in the oral group and 72 (80%) of 90 patients in the intravenous group achieved the primary endpoint (absolute treatment difference 0·5%, 90% CI −9·5 to 10·4). Rates of adverse events were similar, but insomnia was more frequently reported in the oral group (77 [77%]) than in the intravenous group (63 [64%]). Interpretation Oral administration of high-dose methylprednisolone for 3 days was not inferior to intravenous administration for improvement of disability scores 1 month after treatment and had a similar safety profile. This finding could have implications for access to treatment, patient comfort, and cost, but indication should always be properly considered by clinicians. Funding French Health Ministry, Ligue Française contre la SEP, Teva