The study objective is to evaluate the efficacy and accuracy of monitoring prothrombin times at home. It is designed as a randomized, prospective cohort study. Outpatients discharged from a university hospital or a community hospital were used. Fifty patients started on warfarin for the first time who demonstrated an ability to use the monitor and who had not achieved a stable responses to warfarin in the hospital were included in this study. Oral anticoagulation therapy managed using a portable prothrombin time monitor were compared with specialized anticoagulation clinic care. In the 46 patients who completed the 8-week study, the median percentage of time that patients in the home-monitor group (n = 23) were within a range equal to the target prothrombin ratio .+-. 0.3, but always above 1.25, was 93%, compared with 75% for patients in the clinic group (n = 23) (P = 0.003). There was no significant difference between groups in the percentage of time above the therapeutic range; however, the percentage of time that patients were subtherapeutic was significantly greater in the clinic group (P < 0.001). There were no major thromboembolic or hemorrhagic complications in either group. Differences between home monitor measurements and corresponding clinical laboratory measurements using blood samples drawn within 4 hours of the home test were comparable to differences observed between measurements using two different clinical laboratory instruments. The conclusion is that the use of a portable prothrombin time monitor by patients at home is feasible and provides accurate measurements. Patients doing home monitoring achieve superior anticoagulation control compared with those receiving standard anticoagulation clinic care.