-Shock is accompanied by generalised splanchnic hypoperfusion, and splanchnic organs like the pancreas can be damaged, as shown in animal experimental models and in humans, by the presence of high plasma concentrations of trypsin and other pancreatic enzymes. In order to design a radioimmunoassay technique (RIA) for the measurement of equine trypsin-like immunoreactivity (TLI) in biological fluids, trypsin was purified (with purity 96 %) from the equine pancreas by extraction in an acid medium, ammonium sulfate precipitations, gel filtration chromatography and, after activation of trypsinogen into trypsin, affinity chromatography. Gel polyacrylamide electrophoresis showed a monomeric enzyme with a molecular weight of 27 kDa. The purified equine trypsin served for the immunisation of rabbits in order to obtain a specific antiserum, and the labelled antigen was prepared by iodination of equine trypsin with 125 I. The RIA was based on the binding of the antigen to the antibody followed by the separation of the antigen-antibody complex by immunoprecipitation in the presence of sheep anti-rabbit gammaglobulins and the assay of the radioactivity in the precipitate. The RIA showed good sensitivity, specificity, precision, accuracy and reproducibility. The reference mean value of TLI in the plasma of healthy horses (n = 20) was 30.01 ± 6.84 ng/mL (upper confidence limit 50.52 ng/mL; p < 0.01). Three horses with non strangulating intestinal obstruction without shock showed TLI values within normal limits whereas 5 of 7 horses with strangulation obstruction showed TLI levels above the upper confidence limit. Further studies using the RIA and the enzymatic assay should be performed in order to confirm the role of the pancreas in equine intestinal obstruction. equine trypsin / purification / radio-immunoassay / reference range / intestinal obstruction ³