BACKGROUND Corticosteroids are often used to improve the rate of recovery from acute exacerbation in multiple sclerosis (MS) patients. However, it is still unclear just how relatively effective these agents are and the type of drug, optimal dose, frequency, duration of treatment and route of administration are unknown. OBJECTIVE The object of this review was to determine the efficacy and safety of corticosteroids or ACTH in reducing the short and long term morbidity from MS. Moreover, we wished to examine from indirect comparisons if the effect of corticosteroids is different according to different doses and drugs, routes of administration, length of treatment. METHODS A search strategy developed for the Cochrane MS Group (last searched: June 1999) completed with handsearching and personal contacts with trialists and pharmaceutical companies was used. METHODS All randomised, double-blind, unconfounded trials comparing corticosteroids or ACTH to placebo in patients with MS, treated for acute exacerbations, without any age or severity restrictions, were evaluated. METHODS Two reviewers independently selected articles for inclusion, assessed trials' quality and extracted the data. A third reviewer cross-checked them and disagreements were resolved by a joint discussion. RESULTS Six trials contributed to this review; a total of 377 participants (199 treatment, 178 placebo) were randomised. The drugs analysed were methylprednisolone (MP) (four trials, 140 patients) and ACTH (two trials, 237 patients). Overall, MP or ACTH showed a protective effect against the disease getting worse or stable within the first five weeks of treatment (odds ratio[OR]=0.37, 95% confidence interval [CI] 0.24-0.57) with some but non significant greater effect for MP and intravenous administration. Short (five days) or long (15 days) duration of treatment with MP did not show any significant difference. Only one study (with 51 patients) reported data after one year of follow-up: no difference between oral MP and placebo in the prevention of new exacerbations or improvement in long term disability was detected. No data are available beyond one year of follow-up to indicate whether steroids or ACTH have any effect on long-term progression. One study reported that a short term treatment with high dose intravenous MP was not attended by adverse events. On the contrary, gastrointestinal symptoms and psychic disorders were significantly more common in the oral, high-dose MP than in the placebo group. Weight gain and edema were significantly more frequent in the ACTH group than in controls. CONCLUSIONS We found evidence favouring the corticosteroid MP for acute exacerbation in MS patients. Data are insufficient to reliably estimate effect of corticosteroids on prevention of new exacerbations and reduction of long-term disability. Studies assessing long term risk/benefit and adverse effects of corticosteroids in MS patients are urgently needed.