ABSTRACT: INTRODUCTION: Atrial fibrillation affects an estimated 2.5 million Americans and incurs an average annual stroke risk of 4.5% per year. Despite warfarin reducing stroke risk by approximately 66%, prior studies show warfarin usage rates to be about 50%. However, the methods that define warfarin as "inappropriate underutilization" might not be sensitive enough to pick up relative contraindications. We assessed the inappropriate underutilization of warfarin in atrial fibrillation patients at our hospital by abstracting individual patient charts. METHODS: Medical records were reviewed to determine stroke risk factors, warfarin use, and documented contraindications to warfarin use in 364 consecutive patients with atrial fibrillation. RESULTS: Amongst 364 atrial fibrillation patients, 54.6% received warfarin anticoagulation. Overall, 29.5 % of patients had documented reasons for not prescribing warfarin. Primary reasons listed by treating physicians included: gastrointestinal bleed 10.7%, secondary/transient atrial fibrillation 8.2%, and fall risk 6.3%. Only 7.1% of the patients had no documented reasons for the lack of warfarin use. CONCLUSION: Consistent with previous reports, 45.4% of patients in this atrial fibrillation cohort were not prescribed warfarin. However, after reviewing medical charts for documented reasons why warfarin was not used, the inappropriate underutilization rate was only 7.1%. These findings suggest that studies utilizing administrative database and ICD-9 CM coding might overestimate warfarin underutilization.