Summary Background Chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) responds to treatment with corticosteroids, intravenous immunoglobulin, and plasma exchange. We aimed to test whether the standard immunosuppressive drug methotrexate was of use in treatment of CIDP. Methods In a pilot, multicentre, randomised, double-blind, controlled trial we compared oral methotrexate 7·5 mg weekly for 4 weeks, then 10 mg weekly for 4 weeks, and finally 15 mg weekly for 32 weeks (40 weeks' total treatment) with placebo in patients with CIDP requiring intravenous immunoglobulin or corticosteroids. After about 16 weeks, the dose of corticosteroids or intravenous immunoglobulin was decreased by 20% every 4 weeks if participants did not deteriorate. Primary outcome was a greater than 20% reduction in mean weekly dose in the last 4 weeks of the trial compared with the first 4 weeks. Secondary outcomes analysed separately at the mid-trial and final visits measured activity limitations and strength. Analyses were done by intention to treat. This study is registered as an International Standard Randomised Controlled Trial, number ISRCTN73774524. Findings 59 of the 60 enrolled participants completed the trial. 14 (52%) of 27 taking methotrexate and 14 (44%) of 32 taking placebo had a greater than 20% reduction in mean weekly dose of corticosteroids or intravenous immunoglobulin (adjusted odds ratio 1·21, 95% CI 0·40–3·70). There were no clinically and statistically significant differences in secondary outcomes. The one serious adverse event in the placebo group and the three in the methotrexate group were not thought to be related to treatment. Interpretation Oral methotrexate 15 mg weekly showed no significant benefit, but limitations in the trial design and the high rate of response in the placebo group meant that a treatment effect could not be excluded. This study can inform design of future trials in CIDP. Funding The GBS/CIDP Foundation International.