In recent years it has been demonstrated widely that relationships do exist between molecular weight distribution and the non‐Newtonian flow of linear amorphous polymers. Variations in distribution cause marked deviations from general reduced variable curves for non‐Newtonian flow. This conclusion means that reduced variables must involve more than a single molecular weight average. This complexity has been evaluated in this study by equating shear relaxation spectra for polymer solutions to their respective molecular weight distributions. Molecular weight distributions have been generated by column fractionation, measured by gel permeation chromatography, and expressed in terms of their standard deviations. The distributions, expressed in the common form ofMw/Mn,ranged from 3.07 to 1.01. The procedure has been successfully applied to non‐Newtonian flow curves for concentrated solutions of polyisobutenes. This new relationship may be suitable for expressing the non‐Newtonian flow of other linear amorphous polymers. A general approach of this type is of value for predicting and tailoring the flow properties of polymers.