Several models are described for the measurement of drug effects on bleeding. The-haemostatic mechanism in large vessels is largely dissimilar from that in small vessels. Pharmacological bleeding models in which small vessels are used can be expected to be most relevant for the clinical situation, but these models have the disadvantage of a low reproducibility. Therefore, We developed a model in which reproducible bleeding is induced in small vessels. Male rats are anaesthetized with Nembutal®. The bleeding is induced by ripping off a flap of abdominal slcin from the underlying tissue. The “wound” is covered with a gauze and blood loss·is measured. Low doses of heparin (~350 IU/kg i.v.) which block the formation of fibrin do not affect the blood loss. However, higher doses of heparin (~700 and 1400 IU/kg i.v.) increase the blood loss significantly. Blood loss is also increased by treatment withanti-platelet serum. Acetylsalicylic acid, sulfinpyrazone and Org 4122 in doset active in experimental arterial thrombosis, do not affect the blood loss. There is no agreement in literature about the effects of acetylsalicylic acid and sulfinpyrazone in clinical bleeding time tests, but heparin, when given as i.v. bolus injections (7500-10000 IU) induces bleeding in quite a number of patients. Because our results are comparable to results obtained in humans and because of the good reproducibility, this test has advantages above other existing animal bleeding tests.