A method was developed for calculating the molecular weight distribution of a polymer melt from its rheology, specifically dynamic or relaxation moduli data. The molecular weight range covered by the solution is specified by the source data or the entanglement molecular weight, and the blending rule consistent with double reptation was used. The solutions for both the discrete relaxation spectrum and final molecular weight distribution were obtained by Tikhonov regularization with the molecular weight distribution solution being particularly sensitive to the value of the regularization parameter λRused in the calculations. Here, λRwas varied and patterns developed between the most appropriate values for an acceptable solution and the polydispersity of the corresponding samples. For relatively low λRvalues, the algorithm was precise enough to resolve individual components in nearly monodisperse, binary blend, and multimodal systems with as many as 13 components. Results for commercial materials of varying polydispersity generally agreed with the molecular weight distributions from chromatography with slightly higher λRvalues, particularly the location of peak molecular weights. Commercial polyolefins, which required still higher λRvalues, gave consistently poorer results due to restricted molecular weight ranges allowed by the source data.