A double-blind, placebo-controlled, crossover study was designed to compare steroid requirements between placebo and methotrexate (MTX) treatment in subjects with corticosteroid-requiring asthma. Subjects began with a steroid taper and then were randomized to a 3-month trial of drug or placebo therapy. Subjects received 15 mg of MTX a week or identical placebo. A 1-month washout period was completed before the crossover trial. Symptom scores, peak flow rates, spirometry, and beta-agonist frequency were closely monitored. Ten subjects completed the study. The average dose of prednisone during the placebo-treatment period was 11.97 mg/day compared to 8.37 mg/day while subjects were taking MTX. This was a 30% reduction in daily steroid requirement (p less than 0.01). Symptom scores and spirometry did not differ between the crossover trials, and overall clinical status was not altered. Complications from MTX were mild and included anorexia, alopecia, and stomatitis. All complications resolved with dose reduction or when MTX was stopped at the end of the study. No subjects withdrew from the study because of MTX complications. Low-dose MTX significantly reduced the steroid requirement in this group of subjects with steroid-dependent asthma. This reduction in steroid requirement was obtained without altering clinical status and without significant complication.