SAn ageing population and the continuing expansion of clinical indications for coumadin therapy have increased pressure on hospital anticoagulant clinics. One solution is patient self-testing (PST) of the international normalized ratio (INR) using capillary blood samples on point-of-care coagulation monitors at home. We conducted a prospective study to determine whether patients can achieve accurate INR values through PST, using the CoaguChek S (Roche Diagnostics, Lewes, UK). The main outcome measurements were: comparability of INR values obtained by PST and the hospital laboratory, patient acceptability as assessed by a questionnaire and anticoagulant control. Eighty-four patients [53 men, 31 women; median age 59 years (range 26-83)], receiving long-term oral anticoagulation (warfarin), were recruited from our Anticoagulation Clinic. Patients were randomized to weekly self-testing or continuing 4-weekly hospital laboratory monitoring of INR. Comparison of INRs (n = 234) showed no significant differences between the CoaguChek (median INR 3.02) and laboratory testing (median INR 3.07). There was excellent correlation between the two methods (r = 0.95), with 85% of CoaguChek results within 0.5 INR units of the laboratory method. On four occasions, differences of >1 unit INR were obtained, but in each case the patient's anticoagulation was unstable (INR >4.5 by both methods) and the differences in INR would not have altered patient management. 87% of patients found self-testing straightforward, 87% were confident in the result they obtained and 77% preferred self-testing. We conclude that PST is a reliable alternative to hospital clinic attendance and is acceptable to the majority of suitably trained patients.