Equilibrium-dialysis experiments showed that Tris or citrate in the solution prevented copper from occupying completely the specific metal-binding sites on human transferrin. Differential measurements of sedimentation velocity under conditions where two atoms of copper per molecule of protein were bound showed an increase in s(0) (20,w), relative to that of the apoprotein, practically the same as that produced by two atoms of iron. Gel-filtration experiments made under the same conditions to investigate the effect of copper binding on the Stokes radius of the protein showed merely that it lost most of the metal as it passed down the column.