We examined the effect of dietary supplementation of linoleic acid (LA), arachidonic acid (AA) or eicosapentaenoic acid (EPA) to rats fed a diet low in linoleic acid on in vitro and in vivo production of prostacyclin. Male Sprague Dawley rats were fed a high-fat diet (50% energy as fat, 1.5% linoleic acid) for two weeks. Three of the groups were then supplemented orally with either 90 mg/d of LA, AA or EPA, all as the ethyl esters, for a further two weeks while remaining on the high-fat diet. Forty-eight hour urine samples were collected at the end of the second and fourth weeks. In vivo prostacyclin production was determined by a stable isotope dilution, gas chromatography/mass spectrometry assay for the major urinary metabolite of prostacyclins (2,3-dinor-6-keto-PGF1α or PGI2-M and Δ17-2-3-dinor-6-keto-PGF1α or PGI3-M). In vitro prostacyclin production was determined by radioimmunoassay of the stable metabolite (6-keto-PGF1α) following incubation of arterial tissue. Oral supplementation with AA resulted in a rise in plasma and aorta 20:4n-6, and increased in vitro prostacyclin and urinary PGI2-M production. EPA supplementation resulted in a rise in plasma and aorta 20:5n-3 and 22:5n-3, and a decline in plasma 20:4n-6, but not in the aorta. In the EPA-supplemented group, the in vitro prostacyclin and the urinary PGI3-M increased, but urinary PGI2-M decreased. The increase in in vitro prostacyclin production in the EPA-supplemented rats was unexpected and without obvious explanation. Supplementation with LA had minimal effect on fatty acid composition of plasma or aorta and caused no change in prostacyclin production with either method. The in vivo measure of prostacyclin production was positively correlated with aorta AA levels, and negatively correlated with aorta levels of EPA. There was a significant positive correlation between the in vitro production of prostacyclin and the in vivo production (as measured by the urinary prostacyclin metabolite level), despite the differences observed in the EPA-fed group. There was a high inter-animal variability in prostacyclin production using either method. These results indicate that dietary AA stimulates and dietary EPA reduces in vivo PGI2 production in the rat. An equivalent amount of dietary LA was without effect.