The state of knowledge with regard to the static and cyclic liquefaction behavior of sands (sands according to the Unified Soil Classification System) has progressed tremendously in the last twenty years (National Research Council, 1985). In fact, based upon the so-called steady state concepts of Casagrande (1975), Castro (1975) and Poulos (1981), it is generally accepted that the end or steady state condition of a liquefied loose sand is the same whether due to static, dynamic, or cyclic undrained loading. Even so, there is still much work to be done, particularly in regard to the post peak (or strain softening) response that follows the initiation of liquefaction. It is necessary that the post-earthquake stability of earth embankments, in which there is such liquefiable material, be assessed based on the residual strength of that type of material. At present, the two approaches to the evaluation of residual strength are less than adequate for the task. One approach is based on an empirical relationship as derived from a limited number of field case studies involving the failure of embankment slopes via liquefaction. Alternatively, a laboratory approach for evaluating residual strength involves running static consolidated-undrained triaxial tests on both reconstituted and undisturbed samples. Such testing requires a degree of sophistication and experience that is beyond the capability of the great majority of geotechnical firms. It is shown here, however, that the laboratory determined undrained steady state (or residual) strength and the associated critical confining pressure can be assessed in a straightforward manner from drained triaxial tests with volume change measurements based upon a proposed effective stress interpretation of such undrained behavior. Accordingly, it should be possible for all geotechnical firms with the standard capability to perform drained triaxial tests on sands (with associated volume change measurements) to accurately assess the residual strength of such liquefying material. The same procedure also allows the prediction of the whole undrained stress-strain curve and the corresponding effective stress path. (A)