Gluten withdrawal from the diet is occasionally used speculatively in the management of multiple sclerosis. To assess whether there might be any rational basis for such a measure we have undertaken morphological and biochemical studies of the jejunal mucosa in 14 patients with multiple sclerosis. All were found to have morphologically normal villi, and quantitative estimation of surface-to-volume ratios gave values which did not differ from control subjects. Intraepithelial lymphocyte counts were normal. Antigliadin antibody titres were not raised in any patient. Estimation of activity of the brush border disaccharidases (sucrase, lactase, and maltase (showed that the mean level of each enzyme did not differ significantly from control subjects. Analytical subcellular fractionation of the biopsies showed no changes in the distribution or activity of marker enzymes for the brush order, lysosomes, mitochondria, cytosol, peroxisomes, or endoplasmic reticulum. It is concluded that there are no gross morphological or biochemical abnormalities in the jejunal mucosa in patients with multiple sclerosis and, therefore, that the use of gluten-free diets cannot be justified on the assumption that these patients suffer from a coeliac-like lesion of the small intestine.