The causes of red cell damage in artificial prostheses have been studied by several investigators. The effects of shearing stresses on cells have been investigated in the concentric cylinder viscometer. It has been demonstrated that at high shearing rates, such as can occur in an insufficient artificial valve, red cells undergo hemolysis and morphological changes similar to those seen in vivo. There are difficulties involved in the interpretation of studies of blood trauma. The difficulties arise through the possibilities of trauma from secondary effects in addition to the damage caused by shear stress. These secondary effects were studied experimentally in a concentric cylinder viscometer. One important effect is that associated with material interaction. It is well known that cell damage can result from interaction with solid surfaces. Therefore, there is a problem distinguishing damage from surface effects as opposed to damage from any other source. The surface interaction was investigated by varying the surface to volume ratio of the concentric cylinder viscometer by a factor of 2.68. It was found that at high shear stresses, above the critical shearing stress, no effect could be observed over the range of shear stresses investigated. Another secondary effect investigated was that of centrifugal forces. This was done by adding serum albumin to the blood to vary the relative density of the plasma to the cells. This addition of albumin increases the viscosity. Therefore, blood can be subjected to the same shear stresses at lower centrifugal fields. It was found that the centrifugal field had little effect on hemolysis and that in the situation with some of the cells being neutrally buoyant there was a slight augmentation of hemolysis rates. In some concentric cylinder viscometers used for blood studied a buffer layer is provided to separate the blood undergoing shear from a gas interface. There exists the possibility of mixing between the blood undergoing shear and that in the buffer layer. This was investigated using a dual isotope tag. It was found that mixing does occur. A model is provided to account for the mixing and to calculate the true herriolysis rates. It was also found that at high rotational rates hemolysis does occur in the buffer layer. As blood is sensitive to thermal damage the heat dissipation characteristics of the concentric cylinder viscometer were investigated. Design curves were obtained for steady state and for transient operation.