A cohort of patients with an INR >7.0 were identified prospectively and compared with a group of patients with stable anticoagulant control. During the study 15 100 INR measurements were recorded and 31 (0.2%) were >7.0. Odds ratios of patient characteristics were calculated as an estimate of relative risk for the development of a high INR. The highest risk factor was a target INR of 3.5 (OR 7.3, 95% CI 2.6-20.2). The second highest risk factor was antibiotic therapy in the 4 weeks preceding the high INR (OR 6.2, 95% CI 1.4-27.7). Bleeding was reported more frequently in the high INR group (OR 5.4, 95% CI 2.1-13.9). Five major bleeds occurred in this group compared to none in the stable group. This analysis identifies risk factors for over-anticoagulation and hence when to intensify monitoring and when to consider pre-emptive warfarin dose reductions.