Low pressure sensitive Fujifilm was used to measure the load distribution between the resected tibial surface and a tibial component at axial loads up to 3,000 N for a rigid interface, a compliant interface of dacron double-sided velour, and a cemented interface. The pressure patterns consisted of a multitude of small red dots, generally reflecting the slight irregularities of the cut surface and the stiffness of the cancellous bone at the surface. The pressure patterns were photographed with high-contrast film and input into a computer using a photodiode matrix camera. The data were analyzed to yield the number of contact points for each sample. The velour was more effective in distribution of load to the proximal tibia than the rigid and cemented interfaces, while there was no significant difference between the cemented interface and the rigid interface. A second series of tests showed significant increases in contact points from rigid to one layer to two layers of velour. Cyclic axial loading tests were performed to study the characteristics of rigid and compliant interfaces in a model of in vitro subsidence. Static pressure patterns taken at regular intervals showed that subsidence occurred in vitro in up to ! h of the tibias, and that the regions of load transfer could change with time. A model of subsidence was proposed and it was suggested that a velour layer could inhibit the subsidence.