Variability in the control of oral anticoagulant therapy has been associated with a heightened risk of complications. We compared control of anticoagulation between two commonly used coumarins, phenprocoumon and acenocoumarol, and among anticoagulation clinics. All qualifying patients were managed at six regional anticoagulation clinics in the Netherlands. This retrospective cohort study compiled data during a three-year period from a computerised dosing and management system. Anticoagulation control was expressed as the percent of time within the therapeutic range and stability expressed as the time-weighted variance in the international normalised ratio (INR). Data were available for 22,178 patients of whom 72% were treated with acenocoumarol. INRs of patients who received phenprocoumon were within the therapeutic range 50% of the time compared with 43% for acenocoumarol (OR 1.32, 95% CI 1.24-1.41). Moreover, patients on phenprocoumon required 15% fewer monitoring visits and had more stable INR values. These observations were consistent for all six clinics. There were also sizable differences between the clinics with respect to control and stability of anticoagulation that were stable from year-to-year and were unrelated to the drug used. With its longer half-life of three to five days, phenprocoumon produces more stable anticoagulation than acenocoumarol and should generally be the drug of choice when these are the available choices. The differences observed among clinics suggest that certain clinics employ policies and practices resulting in better control of anticoagulation.