In this paper, we reassess the local solute redistribution equation (LSRE) of macrosegregation which, since it first appeared in 1960s, has served as a cornerstone for understanding the composition variations that occur in the solidification of alloys. We highlight some anomalies in earlier literature, in particular as regards the prediction of remelting as a precursor to the formation of channel segregates (freckles, A-segregates and V-segregates) in casting processes. Also, we suggest extensions to the LSRE for situations where solute diffusion in the solid phase is not negligible, as well as when the mode of solidification is unconsolidated equiaxed dendritic, rather than columnar/consolidated equiaxed dendritic. In addition, the significance of the equation for latter-day numerical computations of macrosegregation is also discussed.