The first measurable event upon interaction of artificial surfaces with blood is adsorption of proteins within seconds or minutes. At a later stage, blood cells interact with the surfaces through the initially deposited protein layer. The chemical composition of the surface is only one criterion for differential deposition of various plasma proteins, with molecular motion (polymer chain ends, loops and their flexibility), and topography (roughness, porosity) of the surface decisively influencing the interactions as well. Initially incompatible surfaces, which may be dangerous to the patient, may be rendered compatible by physicochemical surface modifications. Modern methods to estimate biocompatibility have become so sensitive that they may detect biological modifications of the blood after contact with a surface, which has no consequence for the patient. In these cases it is often difficult to decide whether a material should be classified as biocompatible or non-biocompatible. This paper discusses some methods that we have used for the study of biocompatibility of extracorporeal circuitry. It seems to us that minute signs of laboratory evidence for bioincompatibility should not preclude usage of the material in a clinical setting.