Drug–drug interaction (DDI) studies involving warfarin are typically conducted with subtherapeutic doses of warfarin to ensure the safety of volunteers. However, this approach may potentially have a systemic bias of underestimating pharmacodynamic (PD) DDI effect on warfarin at therapeutic levels of anticoagulation. We demonstrate here the utility of model‐based DDI prediction for a clinically relevant warfarin regimen, using the example of epacadostat (INCB024360), the first‐in‐class indoleamine 2,3‐dioxygenase 1 inhibitor in clinical development as a novel orally active immuno‐oncological therapy. Observed data from a dedicated clinical DDI study using subtherapeutic warfarin suggested warfarin pharmacokinetics (PK), but not PD (anticoagulation), was significantly affected by concomitant epacadostat. However, subsequent PK/PD modeling and simulations indicated a clinically important DDI effect on warfarin PD at a higher baseline of the international normalization ratio (INR) and enabled recommendation of warfarin dose adjustment that is dependent on epacadostat dosing regimen and target INR.