A method for the determination of immunoreactive somatostatin in rat plasma is described. Blood specimens were collected into aprotinin and EDTA. Plasma was separated, immediately diluted with acidified acetone and ultrasonicated. The resultant supernatant was lyophilised. The dilution curve of the material thus extracted was parallel to that of synthetic somatostatin. The material was eluted mainly in a similar position to that of synthetic somatostatin on Sephadex G-25(0 column chromatography. The somatostatin immunoreactivity was degraded significantly from the pre-incubated value of 846 + 86 pg/ml (n = 4, mean _+ SEM) to 102 __ 16 pg/ml in the same manner as that of synthetic somatostatin when incubated with one ml of fresh rat plasma at 37 ~ for 30 min. The mean recovery in quadruplicate of immtmoreacfive somatostatin at concentrations of 100, 200 and 400 pg/ml was 83 ___ 7, 95 ___ 4 and 76 ___ 4%, respectively. Using this method, plasma immunoreactive somatostatin responses to arginine, glucose and glucagon infusion were measured in pentobarbital anaesthetized rats. The mean basal plasma immtmoreactive somatostatin concentration in the jugular vein was 35 + 3 pg/ml (n = 7), while that in the hepatic portal vein was 120 + 17 pg/ml (n = 7). Infusion of arginine, glucose and glucagon all resulted in 2-3 fold increases in portal plasma immunoreactive somatostatin concentration.