Mice were fed for 2 mo diets having ratios of alpha-linolenate [18:3 (n-3)] to linoleate [18:2(n-6)] of < 0.01, 0.36, 1.0 and 3.9. Proportions of safflower seed oil and perilla seed oil were adjusted to obtain these ratios. The dietary alpha-linolenate to linoleate balance was reflected in the proportion of (n-3) and (n-6) highly unsaturated fatty acids with 20- and 22-carbon chains in spleen phospholipids, but the ratio did not affect the proportion of T lymphocyte subsets expressing CD4 and CD8 antigens in splenic leukocytes. The immunoglobulin (Ig) G and IgM responses against sheep red blood cells when estimated as plaque-forming cells present in spleen, were not affected significantly by the diets. However, the serum hemagglutinin titer was slightly but significantly higher in the high alpha-linolenate diet group [18:3(n-3)/18:2(n-6) = 3.9] than in the dietary group with 18:3(n-3) to 18:2(n-6) ratios of 0.36 and < 0.01. In contrast, the IgE antibody response against egg albumin, as well as the mortality from anaphylactic shock induced by a second challenge with antigen, was significantly lower in the high alpha-linolenate diet group [18:3(n-3)/18:2(n-6) = 3.9] than in the high linoleate diet [18:3(n-3)/18:2(n-6) < 0.01] group. These results, together with the reported suppressive effects of a high alpha-linolenate diet on the formation of lipid-derived allergic mediators, support the hypothesis that raising the (n-3) to (n-6) ratios of diets would be effective in reducing the severity of immediate-type allergic hypersensitivity.