Spontaneously hypertensive rats (SHR) and normotensive control, Wistar/Kyoto (WKY) rats through two generations were fed a semipurified diet supplemented either with safflower oil (rich in linoleate) or with perilla oil (rich in alpha-linolenate). The cerebral lipid contents and phospholipid compositions did not differ between the two dietary groups of SHR rats. There were also no differences in the unsaturated/saturated ratios of individual phospholipids or the proportions of plasma-logens. However, the proportions of (n-3) and (n-6) fatty acids were significantly different. Decreases in the proportions of docosahexaenoate [22:6 (n-3)] in phosphatidylethanolamine and phosphatidylserine in the safflower oil group were compensated for with increases in the proportions of docosatetraenoic [22:4 (n-6)] and docosapentaenoic [22:5 (n-6)] acids as compared with the perilla oil group. These differences in phospholipid acyl chains were much smaller than the difference in the proportions of linoleate and alpha-linolenate of the diets. In a brightness-discrimination learning test, the total number of responses to the positive and negative stimuli were less in the groups fed perilla oil. However, the alpha-linolenate-deficient group took longer to decrease the frequency of R- responses and therefore longer to learn the discrimination. Consequently, the correct response ratios were higher in the perilla oil groups than in the safflower oil groups. Thus, the dietary alpha-linolenate/linoleate balance influenced the (n-3)/(n-6) balance of polyenoic fatty acids differently among brain phospholipids.(ABSTRACT TRUNCATED AT 250 WORDS)