Thirty years ago we reported that erythrocyte pyridoxal kinase activity of African-Americans was strikingly lower than that of persons with European ancestry in a tissue-specific manner. At the time, it was impossible to elucidate the mechanism by which evolution had selectively lowered the enzyme activity in one cell type but not in others. We have now identified a promoter mutation with potential erythroid-specific properties that could be the basis of a novel mechanism of controlling cell-specific decreased activity of an essential enzyme.