Apparent kinetic parameters have been measured for the transfer of N-acetyl-D-neuraminic acid (Neu5Ac) from CMP-Neu5Ac to analogues of the Gal(beta 1-4)GlcNAc (type II) and Gal(beta 1-3)GlcNAc (type I) substrates by the rat liver Gal(beta 1-4)GlcNAc alpha 2,6-sialyltransferase and the Gal(beta 1-3/4)GlcNAc alpha 2,3-sialyltransferase. In these acceptor analogues, the substituents of the pyranose rings were modified, particularly by deoxygenation, to identify (i) the key polar groups required for efficient transfer and (ii) the substituents that can be removed or modified. A topography including the 6-hydroxyl of the beta Gal and the 2-acetamido of the GlcNAc unit is required for transfer to a terminal type II disaccharide by the alpha 2,6-sialyltransferase. The other hydroxyls can be replaced by hydrogen without a substantial decrease in activity. The alpha 2,3-sialyltransferase requires the 3-, 4-, and 6-hydroxyls of the terminal beta Gal and some contribution from the subterminal sugar. This may explain the cross-reactivity of this enzyme for the type I and type II acceptors. For both enzymes, an influence of the hydrophobic nature of the aglycone is noticed. The results allow an evaluation of the efficiency of the transfer of Neu5Ac to modified substrates.