While most structural studies of the Na,K-ATPase support a subunit stoichiometry of one α-subunit to one β-subunit, the exact quaternary structure of the Na,K-ATPase and its relevance to enzyme function is the subject of much debate. Formation of a higher order enzyme complex is supported by our previous study demonstrating specific α/α interactions among the rat Na,K-ATPase isoforms (α1, α2, α3), expressed in virally infected Sf-9 insect cells and among native α isoforms in rat brain (1). This detergent-resistant association was not observed in insect cells coexpressing the homologous gastric H,K-ATPase α-subunit, nor was it dependent on the coexpression of the β-subunit. To delineate domains necessary for α/α assembly, a series of H,K-ATPase. Na,K-ATPase chimeras were constructed by combining the N-terminal, cytoplasmic midregion and C- terminal segments derived from the Na,K-ATPase (N) and the H,K-ATPase (H) α- polypeptides (HNN, HNH, NHH, NHN, and HHN). The α-subunit chimeras were coexpressed with the Na,K-ATPase α1-subunit in Sf-9 cells using the baculovirus expression system. Specific and detergent-stable association is observed between the Na,K-ATPase α-subunit and the HNN and HNH chimeras, but not with the NHH, NHN, or HHN chimeras. Consistent with the Na,K-ATPase cytoplasmic domain as being necessary for α/α interactions, the full- length α-subunit stably associates with an α N-terminal deletion mutant (ΔGly2-Leu273), but not with an α cytoplasmic deletion mutant (ΔArg350-Pro785). In addition, the naturally occurring C-terminal truncated α1 isoform, α1T (ΔGly554 to C terminus), does not associate with the α1-subunit in Sf-9 cells coexpressing both polypeptides. Thus, a cytoplasmic region in the α-subunit (Gly554-Pro785) is necessary for specific α/α association. The same cytoplasmic region contains a strongly hydrophobic segment that, by analogy with oligomerization of water-soluble proteins, may form the interface of the extramembranous α/α contact site.