It has been established that herbal intake affects the anticoagulation effects of warfarin, but the long-term impact on anticoagulation control is unclear. We sought to investigate the effect of concomitant herbal intake on anticoagulation control in patients with nonvalvular atrial fibrillation (AF) treated with warfarin. The effects of common herbs were determined by monitoring the international normalized ratio in 250 patients with AF (69 ± 10 years, 50% male). All the patients had been prescribed warfarin therapy for at least 6 months before enrollment, and their dietary intake, including the type and the frequency of common herbs, was recorded using a standardized questionnaire. Up to 50% of the patients reported consumption of foods with herbal ingredients, including garlic (80.4%), ginger (74.8%), green tea (50.4%), and papaya (55.2%) but rarely herbal drugs such as danshen (1.2%), dong guai (0.8%), fenugreek (1.2%), psyllium seed (0.4%), and ginseng (4%). Infrequent users (1 kind of herb for <4 times per week and nonusers) were more likely to have an international normalized ratio within the optimal therapeutic range (2.0-3.0) than frequent users (>1 kind of herb for ≥4 times per week) (58.1% vs 51.1%, P = 0.046). In conclusion, the patients with AF treated with warfarin had little knowledge about the potential interaction of herbal substances in foods with warfarin. The patients who consumed common herbs at least 4 times per week had suboptimal anticoagulation control with warfarin.