Experimental and theoretical methods were used to investigate the linearity of the stress response of articular cartilage to ramp and sinusoidal tests in confined compression, as well as the role of cartilage-porous platen and lateral confinement boundary conditions in determining material responses. With respect to linearity, we posed the question as to whether the elicited stress responses to ramp compression, ramp release and sinusoidal tests were similar. With respect to boundary conditions we inquired as to the necessity of specifying a detailed interdigitating contact with the porous filter and of specifying the level of confinement present at the lateral edge of the disk. We found that the stress responses to the three types of tests were dissimilar, with ramp compression the only test exhibiting linear behavior. Ramp release from a static compression offset was non-linear in a manner such that the cartilage maintained a compressive stress higher than expected by a linear theory. Sinusoidal compression also displayed a non-linear response consistent with the presence of a release phase in each cycle. The actual boundary conditions present at the cartilage/porous-filter interface were visualized histologically. Areas (tens of microns) of cartilage in contact with the metal of the filter were interspersed with areas expanded into the pores of the filter. Finite-element analysis incorporating this information suggested that precise specification of this interface and of the level of the extent of lateral confinement would be necessary for the estimation of material properties, such as the hydraulic permeability, from these tests. The trends of the linearity studies did not appear to be significantly affected by the problems posed by these difficult to quantitate boundary conditions. The non-linear cartilage response to release and sinusoidal displacements therefore appear to be physiologically interesting. The maintained, that is higher than would be linear, compressive stress observed during release may be a beneficial adaptation to repeated loading where temporal variations in tissue stresses would be minimized.