Abstract Inelastic deformation in metals is known to be a time-dependent phenomenon; however, the theoretical relations between load and deformation do not, in general, include the effect of time. Most of the experimental data, confirming these theoretical relations, have been obtained using screw-powered or hydraulic testing machines in which the load “falls off” as inelastic deformation occurs. Furthermore, these tests have been made without regard to the length of time necessary for inelastic deformation to cease at any given load. This investigation was undertaken to determine the effect of testing members under dead load where each increment of load could be maintained on the member a sufficient length of time until equilibrium was attained. Five statically determinate beams, four mild-steel beams of various cross sections, and one rail-steel rectangular beam were tested in a dead-load testing machine. The load-deformation diagrams were found to have a stairstep appearance, unlike the smooth continuous curves usually reported. The ultimate load-carrying capacity of the mild-steel beams was found to be considerably below that predicted by theory, and time delays of appreciable magnitude were observed for the initiation of inelastic deformation at a given load. A more stable behavior was observed for the rail-steel beam under dead load, and the experimental results were found to correlate well with theory.