The effect of dietary alpha-linolenic acid (18:3n-3) and its ratio to linoleic acid (18:2n-6) on platelet and plasma phospholipid (PL) fatty acid patterns and prostanoid production were studied in normolipidemic men. The study consisted of two 42-d phases. Each was divided into a 6-d pre-experimental period, during which a mixed fat diet was fed, and two-18 d experimental periods, during which a mixture of sunflower and olive oil [low 18:3n-3 content, high 18:2/18:3 ratio (LO-HI diet)], soybean oil (intermediate 18:3n-3 content, intermediate 18:2/18:3 ratio), canola oil (intermediate 18:3n-3 content, low 18:2/18:3 ratio) and a mixture of sunflower, olive and flax oil [high 18:3n-3 content, low 18:2/18:3 ratio (HI-LO diet)] provided 77% of the fat (26% of the energy) in the diet. The 18:3n-3 content and the 18:2/18:3 ratio of the experimental diets were: 0.8%, 27.4; 6.5%, 6.9; 6.6%, 3.0; and 13.4%, 2.7, respectively. There were appreciable differences in the fatty acid composition of platelet and plasma PLs. Nevertheless, 18:1n-9, 18:2n-6 and 18:3n-3 levels in PL reflected the fatty acid composition of the diets, although very little 18:3n-3 was incorporated into PL. Both the level of 18:3n-3 in the diet and the 18:2/18:3 ratio were important in influencing the levels of longer chain n-3 fatty acid, especially 20:5n-3, in platelet and plasma PL. Production of 6-keto-PGF1 alpha was significantly (P < 0.05) higher following the HI-LO diet than the LO-HI diet although dietary fat source had no effect on bleeding time or thromboxane B2 production.(ABSTRACT TRUNCATED AT 250 WORDS)