The T-cell antigen receptor (alpha beta or gamma delta TCR) is known to associate with four polypeptides (CD3 gamma, delta, epsilon and zeta) to form the TCR-CD3 complex. Although the six chains are well characterized, the molecular mass of the TCR-CD3 complex and stoichiometry of the components are currently uncertain. We analysed the TCR of a T-T hybridoma which expresses two distinct heterodimers. When the hybridoma was incubated with a mAb (MR9.2) specific for the V alpha(10)V beta(5.1) heterodimer, both of the heterodimers were lost from the cell surface, as measured with mAb MR9.2 and MR9.7 (V alpha(1)V beta(1)-specific). The ability to co-modulate V alpha(1)V beta(1) and v alpha(10)V beta(5.1) suggested that TCR complexes could contain two alpha beta-heterodimers. Density gradient sedimentation analysis provided further evidence for higher order TCR. The sedimentation patterns of the TCR were compared to that of the B-cell antigen receptor and the well-characterized VSV membrane G-protein as well as to soluble marker proteins. Maximal cell surface murine and human TCR sedimentation coefficients were substantially greater than the 9-10S predicted for a 210 kDa monovalent alpha beta gamma delta epsilon(2) zeta(2) structure. The TCR sedimented in mild non-ionic detergents as large 18 +/- 3S complexes co-migrating with a 443 kDa marker protein. In contrast, the IgM B-cell antigen receptor had a maximal sedimentation coefficient of 10 +/- 3S, consistent with a predicted size of similar to 300 kDa. Taken together, the results suggested that T-cell antigen receptors can contain more than one alpha beta-heterodimer which could be incorporated into a minimal divalent 10-chain TCR-CD3 complex (e.g. alpha beta gamma epsilon epsilon delta zeta zeta alpha beta).