Side-effects of iohexol myelography were compared after pre-treatment with oral dexamethasone (n = 42) or placebo (n = 44) in a prospective, randomized, double blind, controlled clinical trial. Although myelogram side-effects were commoner in the placebo group, the differences generally did not reach statistical significance. The routine prophylactic use of oral corticosteroids for myelogram side-effects therefore cannot be recommended and our results may cast doubt on their use in the treatment of these symptoms once they have developed. These findings also suggest that inflammatory processes (allergic or chemical irritant), relating to the contrast agent itself, against which corticosteroids might have been expected to act, have at most a minor pathogenetic role in postmyelogram symptomatology.