The control achieved in two anticoagulant clinics over 1 year was studied. The result of 430 patient years of treatment in 732 patients was assessed. Overall, the patients were maintained in the therapeutic range (British ratio 2.0–4.0) 85% of the time, ‘under-anticoagulated’ 10% and ‘overtreated’ 5% of the time. Patients on long-term treatment had better control than those on short-term treatment (87 and 72% of time in therapeutic range, respectively). Short-term patients were ‘undertreated’ one quarter of the time. Assessment of the percentage of time individual patients spent within the therapeutic range was a useful index: 77% short-term patients and 99% long-term patients were controlled more than half the time; and 30% short-term plus 40% long-term patients were controlled all the time. Two complementary methods of assessing therapeutic control are used. The standard of control compares favourably with other reports, but shows areas where improvements can be made. For assessment of clinical benefit from anticoagulants, close quality control of treatment by such methods is essential.