Twenty-one healthy greyhounds with no history or clinical signs of bleeding disorders, and no abnormalities on physical examination, complete blood count, serum biochemistry profiles (in dogs more than five years of age), and SNAP-4DX test for vector borne diseases underwent routine gonadectomies at the Ohio State University Veterinary Teaching Hospital. Blood samples were collected 24 hours before and after surgery by jugular venepuncture for thromboelastography and haemostasis assays (prothrombin time [PT], activated partial thromboplastin time [aPTT], fibrinogen concentration). The magnitude of the bleeding in each patient was estimated using a bleeding scoring system recently validated in greyhounds. Eight dogs were classified as bleeders and 13 as non-bleeders. Thromboelastograph (TEG) tracings in bleeders were different to that of non-bleeders. Neither sex (odds ratio [OR]: 0.148, P=0.05), haematocrit (OR: 0.907, P=0.39), platelet count (OR: 0.996, P=0.65) or age (OR: 0.949, P=0.83) were predictors of the outcome. None of the variables that evaluated clot kinetics, and fibrinolysis (that is, aPTT OR: 0.781, P=0.51; PT OR: 1.337, P=0.63; TEG(R) OR: 1.269, P=0.06; TEG(K) OR: 1.696, P=0.05; TEG(LY60) OR: 1.028, P=0.81) were able to predict the bleeding episodes. Only the TEG variables that represent the fibrin cross-linking of the clot (TEG(angle) OR: 0.903, P=0.03); and the strength of the clot (TEG(MA) OR: 0.833, P=0.03) were considered predictors of the outcome.