Summary The possibility that the haemagglutinin of encephalomyocarditis (EMC) virus might be a glycoprotein was investigated by looking for sugars and amino sugars in purified EMC virus particles grown in Krebs ascites tumour cells in vitro in the presence of isotopically labelled glucosamine, acetyl glucosamine, galactosamine, mannosamine or fucose. With either isotopically labelled glucosamine or galactosamine the virus became significantly radioactive, whereas radioactivity was low in virus grown in the presence of mannosamine, or fucose, possibly because the latter was poorly taken up by the cells; acetyl glucosamine was completely excluded by Krebs cells, so its ability to be incorporated into virus could not be tested. Radioactivity derived from glucosamine remained firmly attached to the purified virus through both rate-zonal and equilibrium density centrifugation and chromatography on calcium phosphate. This radioactivity did not appear due to simple adsorption of either glucosamine or (after synthesis) acetyl glucosamine, or to contamination of the preparation with infected host cell membranes. Estimates of the number of glucosamine molecules incorporated into the virus had values which varied with the position of the isotope in the glucosamine added, suggesting that it was not incorporated unchanged. Polyacrylamide gel electrophoresis of disrupted whole virus showed radioactivity derived from glucosamine to be present in all virus polypeptides, indicating that glucosamine was perhaps metabolized, amongst other things, to amino acids. Disruption of the virus with phenol revealed that 70% of the radioactivity was in the virus RNA, mostly in ribose, the remaining 30% being in virus protein, presumably as amino acids. The total amount of radioactivity that could be in the form of glucosamine did not allow for more than 0.6 glucosamine residues per virus particle. From this it was concluded that glucosamine is not a virus constituent per se, that the EMC virus particle does not contain glycoprotein and, consequently, that the virus haemagglutinin is not a glycoprotein.