Hepatitis B virus surface antigen (HBsAg) could be studied until recently only by isolating it from the blood of carriers, thus making incorporation of radioactive precursors into this protein(s) impossible. The isolation of a cell line producing HBsAg [Alexander et al, 1978] has eliminated this obstacle. The cell line was therefore used for labeling HBsAg either with 35S-methionine or with 35S-cystine. HBsAg was purified by pelleting the component and by isopycnic centrifugation in CsCl gradients. HBsAg-positive fractions (as determined by solid-phase radioimmunoassay) were isolated from the gradients and analyzed in sodium dodecyl sulfate-containing polyacrylamide gels. It was found that although HBsAg contains substantial amounts of 35S-cystine, very little 35S-methionine was incorporated into this protein. In contrast, both labels were found in other structures having a buoyant density of about 1.3 gm/cm3 in CsCl. It was concluded that HBsAg is very low in methionine, and therefore this amino acid should not be used for labeling HBsAg in cells or in a cell-free system. Analysis of 35S-cystine-labeled HBsAg-positive material (buoyant density about 1.2 gm/cm3 in CsCl) revealed five proteins with molecular weights in the range of 48,000-82,000.