An assessment of the experimental findings leads to the conclusion that optimal structural superplasticity results from grain/interphase boundary sliding—diffusion coupled flow. An analysis of the boundary sliding process is presented first. By suggesting that both regions I and IIa (lower stress range of region II) of superplastic flow result from sliding—diffusion coupled flow, and—treating mesoscopic (cooperative) boundary sliding as the rate controlling mechanism for optimal superplasticity, the stress, temperature, and grain size dependences of the strain rate of deformation are predicted. The above equation is then related to the stress exponent n (the inverse of the strain rate sensitivity index m). An analysis for determining the true activation energy for the rate controlling process is presented. Expressions for the distribution of internal stresses arising from sliding and the boundary viscosity are derived, and the presence of an initial unsteady region, predicted in an earlier analysis, is shown to be a natural consequence of this approach.MST/3074