Prostacyclin release from endothelial cells in culture appears increased by the presence of plasma, but occurrence of a similar phenomenon in intact vessels has not been established. In the present investigation release of 6-keto-PGF1 alpha, the stable breakdown product of prostacyclin, was quantitated from canine veins perfused ex vivo for 15 minute periods, using three different perfusates: 1) Hank's balanced salt solution (HBSS), 2) 20% platelet poor plasma (PPP) derived from heparinized blood, in HBSS, and 3) 20% PPP in HBSS with added arachidonic acid (AA). No differences in initial 6-keto-PGF1 alpha release existed among the three perfusates. However, PPP and PPP + AA solutions were associated with lesser declines in release, than occurred with HBSS alone, these differences being statistically significant after 60 min of perfusion (p less than 0.05). When PPP derived from heparinized and citrated blood rather than from only heparinized blood was used, there was a significantly lower release of prostacyclin (p less than 0.05). The latter may be due to citrate binding of calcium. These data indicate that autologous plasma does not alter initial prostacyclin release from freshly harvested canine veins, but that it weakens the decline in release usually following prolonged periods of perfusion.