The most common system for synthesis of cell surface polysaccharides is the Wzx/Wzy-dependent pathway, which involves synthesis, on the cytoplasmic face of the cell membrane, of repeat units, which are then translocated to the periplasmic face by a Wzx translocase and then polymerized by Wzy to generate the polysaccharide. One such polysaccharide is O antigen, which is incorporated into lipopolysaccharide (LPS). The O antigen is extremely variable, with over 186 forms in Escherichia coli. Wzx proteins are also very diverse, but they have been thought to be specific only for the first sugar of the repeat units. However, recent studies demonstrated examples in which Wzx translocases have considerable preference for their native repeat unit, showing that specificity can extend well beyond the first sugar. These results appear to be in conflict with the early conclusions, but they involved specificity for side branch residues and could be a special case. Here we take six Wzx translocases that were critical in the earlier studies on the importance of the first sugar and assess their ability to translocate the Escherichia coli O16 and O111 repeat units. We use gene replacements to optimize maintenance of expression level and show that under these conditions the native translocases are the most effective for their native repeat unit, being, respectively, 64-fold and 4-fold more effective than the next best. We conclude that Wzx translocases are commonly adapted to their native repeat unit, which provides an explanation for the great diversity of wzx genes.