To examine the cost and cost-effectiveness of the use of point-of-care (POC) devices by the general practitioner (GP), in anticoagulation clinic or by the patient in self-testing (PST) and self-management (PSM), compared with standard laboratory testing to realize international normalized ratio tests for patients on long term anticoagulation therapy. An economic evaluation was performed from the Belgian health care payer's perspective using a Markov model. Outcomes data were derived from a meta-analysis and cost data were derived from claims databases. Several scenarios were tested based on number of tests and GP's contacts and probabilistic sensitivity analysis was used to handle uncertainty. Evidence on the impact of POC on mortality was only found for PSM. Therefore, a cost-effectiveness analysis was performed for PSM and for other strategies, only a cost comparison was done. With an unchanged number of tests, POC is cost-saving compared to laboratory testing (probability > 70%). In scenarios where POC induces more tests, results were different: with 52 tests/year, only PSM kept a probability of remaining cost-saving superior to 50%. Except in the case of 100% of GP consultations maintained and 52 tests/year performed, PSM resulted in significantly more "life years gained" (LYG) than usual care and was on average cost-saving. The organisation of long term oral anticoagulation monitoring should be directed towards PSM and, to a lesser extent, PST for selected and trained patients.