In assessing the optimal intensity of anticoagulant therapy, the International Normalized Ratio (INR) at admission is used as a basis for INR-specific incidence rates. In 47 patients suffering a haemorrhagic stroke we tested the assumption that the INR at admission is an acceptable measure for the INR that preceded the haemorrhage. We found high D-dimer levels in 70% of the patients, which indicated activated coagulation and fibrinolysis. This was not of such an extent that it could also be measured with other routine coagulation tests, with the possible exception of two patients. We found normal INRs in 33 non-anticoagulated patients, and only a mildly prolonged INR of 1.9 in one patient, which was most probably caused by a vitamin K deficiency. We concluded that the INR at admission can be used in studies to assess the optimal level of anticoagulation.