Constitutive equations of the so‐called rate‐dependent type were found to be in disagreement with the experimental data of stress development after onset of steady shear flow, observed with a Weissenberg Rheogoniometer. A discussion on the reason for the disagreement is given. It is reported that such transient phenomena can be quantitatively explained if we assume that a transition from the equilibrium entangled structure to the steady structure at the shear rate occurs when the strain in the material reaches a critical value.