The functional relationship between various difinitions of effective stress, i.e., conventional, Biot-Willis-Nur-Byerlee, and theory of interacting continua (Tinc), is explored both theoretically and experimentally. Stress-strain data on dry and saturated Weber sandstone demonstrate that the conventional effective stress law grossly overestimates the pore pressure effect, whereas Biot-Willis-Nur-Byerlee and Tinc laws somewhat underestimate this effect. Failure is examined from a microscopic point of view. It is shown that Tinc-type effective stress laws cannot account for the effect of pore pressure on failure in a simple manner. Failure is one aspect of material response in which the conventional effective stress law is useful. In summary it appears necessary to have two effective stress laws, i.e., one to describe the stress-strain response and a second to delineate the failure surface.