A review of data from creep and work hardening at constant rate is presented, including own data of the minimal creep rate of Cu with coarse grains of about 1 mm size. It shows that the "stationary" (steady state) deformation resistance is ill defined. In particular, deformation instability by fracture causes scatter. The variations of the seemingly stationary deformation resistances are interpreted as a consequence of gradual, continuous changes of the grain structure with strain, in particular the boundary misorientations. The interpretation is supported by results for the deformation resistance after severe predeformation by equal channel angular pressing.