In order to determine the arterio-venous difference in circulating prostaglandin levels the isotope derivative technique was applied to rat plasma for measurements of prostaglandin E2 (PGE2). No arterio-venous difference was observed if the blood was collected without precautions to avoid in vitro prostaglandin production in the sampling tube. In contrast, inhibition of the in vitro prostaglandin release by the infusion of indomethacin into the catheter from which the blood was drawn, showed PGE2 levels in the right atrium of 60 +/- 20 pg/ml (mean +/-SD; N = 8), and peripheral levels insignificantly different from zero. A significant difference in PGE2 levels of right atrial plasma from fed and fasted rats was observed without the use of in vitro indomethacin in blockade (210 +/- 70 pg/ml; N=4, and 330 +/- 35 pg/ml; N = 4, respectively). This difference could not be reproduced if indomethacin was infused into the catheter (60 +/- 20 pg/ml; N = 8 and 90 +/- 35 pg/ml; N = 5, for fed and fasted rats, respectively). Measurements by radio-gas-liquid-chromatography of the arachidonic acid plasma level showed values of 6.8 +/- 1.6 micrometers (N = 7) and 12.0 +/- 3.4 micrometers (N = 7) for fed and fasted rats, respectively. It is suggested that the higher level of PGE2 (without indomethacin) in fasted rats reflects that of arachidonic acid. The present available methods for the determination of primary prostaglandins and prostaglandin metabolites are discussed with respect to their value not only in physiological studies on fluid and electrolyte transport in the gut, but also for clinical use.