The objective was to retrospectively study the initiation of anticoagulant therapy in inpatients of the two major teaching hospitals in Tasmania, Australia. The medical records of a random sample of patients with an admission diagnosis of deep venous thrombosis (DVT) or pulmonary embolism (PE) during the period February 1992 to June 1994 were studied, to examine therapeutic issues including (i) the time taken after commencing heparin to achieve a therapeutic activated partial thromboplastin time (APTT), (ii) when warfarin was commenced, (iii) the time taken after commencing warfarin to achieve a therapeutic International Normalized Ratio (INR), and (iv) the degree of anticoagulant control at the time of discharge from hospital. The medical records of 99 patients (median age: 65 years and range: 16–93 years; 52 females) were studied. Heparin was generally commenced within 4 h of admission to hospital. The median duration of heparin therapy was 5 days (range: 2–26 days). The median number of APTTs performed per patient was 6 (range: 1–24), with most results (60%) being below the optimum range. Warfarin was commenced from day 1 of hospitalization in only 34% of patients. The INR was within the therapeutic range in only 29% of cases when heparin was ceased. The median time taken to achieve a therapeutic INR after starting warfarin was 3 days (range: 1–15 days). The median number of INRs performed per patient was 5 (range: 1–29), with most results (51%) being below the optimum range. The INR was within the therapeutic range in only 55% of cases at discharge. An adverse outcome (bleeding or further clotting) was experienced by 19% of patients. This study has revealed considerable scope for improvement in some aspects of the initiation of anticoagulant therapy.