The presence of glycosyltransferases on surfaces of mammalian cells has been reported by many investigators and a biological role for these enzymes in cell adhesion and cell recognition has been postulated. Critical analysis, however, showed 2 major complications regarding the assay for cell surface glycosyltransferases: 1) hydrolysis of the nucleotide sugar by cell surface enzymes and subsequent intracellular use of the free sugar and 2) loss of cell integrity if trypsinized or EDTA-treated cells were used in suspension assays. We have assayed intact, viable cells in monolayer for cell surface glycosyltransferases using conditions under which intracellular utilization of free sugars generated by hydrolysis of the nucleotide sugar was prevented. Our data demonstrate that the presence of galactosyltransferases on the surface of a variety of cells, including established (normal and virally transformed) as well as nonestablished cells, is unlikely. No evidence for the existence of cell surface fucosyl- and sialytransferases could be obtained, but our data do not exclude the possibility that low levels of these enzymes are present.