The vascular leakage induced by histamine, bradykinin, serotonin and prostaglandin E1and E2was assessed. The test agents were injected intradermally into the shaved thoracic skin of horses and the vascular leakage estimated either semi‐quantitatively by recording the diameter of the lesions or by measuring the actual volume of extravasated plasma in microlitres using iodine‐125‐labelled human serum albumin (125I‐HSA) as a marker in the blood plasma. Using the latter method, the vascular leakage induced by carrageenin and the effect of coadministered prostaglandins E1and E2upon the vascular leakage of both histamine and bradykinin were also investigated. No obvious lesions resulted when serotonin (10‐2mol/1) was injected but histamine and bradykinin produced circular lesions which increased in diameter for approximately 30 min. The size of the lesions and volume of extravasated plasma was dose dependent. On a molar basis, bradykinin (10‐6mol/1, 10‐5mol/1) was more potent than histamine but they were equipotent at 10‐4mol/1. The size of the lesions induced by carrageenin were independent of their anatomical location on the thorax. Except for the second hour, the hourly volume of vascular leakage increased until the fifth hour when the experiment was concluded. The maximum vascular leakage resulting from the injection of prostaglandin E1or E2(1, 10, 100 or 1000 ng) was 7 μ but when co‐administered with bradykinin (10‐6mol/1), the volume of leaked plasma increased from 29 to 78 μ. No synergy was observed when either prostaglandin was co‐administered wit