The rejection caused by the presence of Galalpha1,3Gal (Gal) on the pig vascular endothelium and of natural anti-Gal antibodies in human blood has recently been prevented by the breeding of pigs that do not express Gal, achieved by knocking out the gene for the enzyme, alpha1,3-galactosyltransferase. However, prior to the introduction of nuclear transfer/embryo transfer techniques, a major effort was directed towards reducing Gal expression on pig cells by other methods, such as by cleaving Gal from the underlying substrate, or replacing Gal with an alternative, innocuous oligosaccharide by a process that has been termed 'competitive glycosylation'. Gal has been cleaved by alpha-galactosidase or endo-beta-galactosidase C. Competitive glycosylation has largely targeted replacement of Gal by insertion of a gene for a fucosyltransferase or a sialyltransferase, or by insertions of the gene for N-acetylglucosaminyltransferase III to reduce cell-surface expression of several oligosaccharides. The results of these approaches to render the pig cells less immunogenic to the human immune system are summarized. With regard to the problem provided by Gal expression, the above approaches may be considered by some to be largely obsolete, but the principles underlying them may prove valuable when other antigen targets for human antibodies are definitively identified, if these prove to be carbohydrates.