Prophylactic fresh-frozen plasma (FFP) transfusion is often undertaken in hemodynamically stable patients with a minimally elevated international normalized ratio (INR) prior to invasive procedures, despite little evidence in support of this practice. The authors review the current literature in an attempt to clarify best clinical practice with regard to this issue. Although the activated partial thromboplastin time and prothrombin time-INR are useful laboratory tests to measure specific clotting factors in the coagulation cascade, in the absence of active bleeding or a preexisting coagulopathy, their utility as predictors of overall bleeding risk is limited. Several studies have shown an imperfect correlation between mild elevations in the INR and subsequent bleeding tendency. Furthermore, FFP transfusion is not always sufficient to achieve normal INR values in patients who have mild elevations (< 2) to begin with. Finally, there are risks associated with FFP transfusion, including potential transfusion-associated [disease] exposures as well as the time delay imposed by laboratory testing and transfusion administration prior to initiation of procedures. The authors propose that the current concept of a "normal" INR value warrants redefinition to make it a more meaningful clinical tool. Based on their review of the literature, the authors suggest that in a hemodynamically stable patient population there is a range of mildly prolonged INR values for which FFP transfusion is not beneficial, and is potentially harmful.