Are truly inert biomaterials feasible? Recent mathematical models of coagulation which are reviewed here suggest that such materials are impossible. This conclusion, which is certainly consistent with our collective experimental evidence, arises from the calculation that conversion of Factor XI to XIa never drops to zero even at the highest flow rates and with virtually no Factor XIIa bound to a surface. Residual amounts of XIa are still formed which can in principle kick-off the coagulation cascade. Furthermore, if the flow rates and corresponding mass transfer coefficients are low and in spite of these near-vanishing levels of the initiating coagulants, the surprising result is that substantial amounts of thrombin are produced. On the contrary, under slightly higher flow conditions, there can be more substantial levels of initiating coagulants, yet paradoxically thrombin production is near zero. This article presents a theoretical understanding of the events which take place during the interaction of biomaterials with flowing blood. We follow these events from the time of first contact to the final production of thrombin. The effect of flow and surface activity on the contact phase reactions is examined in detail and the two are found to be intertwined. The common pathway is also examined and here the main feature is the existence of three flow dependent regions which produce either high or very low levels of thrombin, as well as multiple thrombin steady states. In a final analysis we link the two segments of the cascade and consider the events which result. In addition, we note that multiple steady states arise only in the presence of two (thrombin) feedback loops. Single loops or the bare cascade will produce only single steady states. With some imagination one can attribute to the feedback loops the role of providing the cascade with a mechanism to produce high thrombin levels in case of acute need (e.g. bleeding) or to allow levels to subside to 'stand-by' when there is no need for clotting. We present this as a partial answer to the question: Why is the coagulation cascade so complex and what is the importance of the feedback loops?