Objectives : Nowadays the combined use of oriental herbal medicines and western biomedical medicines has been prevalent but controversial. Warfarin has been much reported to interact with some herbal medicines so that it influences prothrombin time(PT) & international normalized ratio(INR). This study was aimed to examine how much warfarin interacts with herbal medicines during treatment of stroke patients Methods : This was a retrospective case control study of 53 patients whowere treated with concomitant treatment of herbal medicines & warfarin. They were within normal limit in liver function, renal function, hematocrit, hemoglobin, and platelet count at first admission lab. We classified them into 2 classes: study group (taking herbal medicines including Panax ginseng, Angelica sinensis, Zingiber officinale, Salvia miltiorrhiza that were reported to interact with warfarin to impact PT (INR) and control group (taking other herbal medicines). We followed up PT (INR) at 5-10 days interval with AST, ALT, BUN, creatinine, hematocrit, hemoglobin, and platelet count. Results : AST, BUN, creatinine, hemoglobin, hematocrit, and platelet count were not changed significantly between first and final tests during the admission period. Only ALT decreased significantly in the control group. Neither baseline nor peak PT (INR) was significantly different between the groups. However, only warfarin dose was significantly correlated with PT and INR (r=0.810, r=0.798, p<0.01). Conclusions : It was concluded that PT(INR) was not influenced with herbal medicines and warfarin but by far dependent on warfarin dose in stroke patients restricted with normal liver function, renal function, and hematocrit, hemoglobin, and platelet count. Further prospective study is needed on larger samples to conclude that the combined therapy of herbal medicines and warfarin is safe.