Functional and structural similarities among a wide variety of endomembrane H+-ATPases suggest that they form a distinct class with a common origin. Immunological studies (Manolson, M. F., Percy, J. M., Apps, D. K., Xie, X. S., Stone, D. K., and Poole, R. J. (1987) in Proceedings of the Membrane Protein Symposium (Goheen, S. C., ed) pp. 427-434, Bio-Rad, Richmond, CA, and M. F. Manolson, J. M. Percy, D. K. Apps, X. S. Xie, D. K. Stone, M. Harrison, D. J. Clarke, R. J. Poole, unpublished data) support this idea and suggest an evolutionary relationship between the endomembrane and F0F1 ATPases. Further examination of relationships necessitates comparison of protein/nucleic acid sequence data. To this end, we have cloned and sequenced the cDNA encoding the 57-kDa polypeptide of the Arabidopsis vacuolar membrane H+-ATPase. To our knowledge, this is the first report of the sequence of a "57-kDa" subunit for plant or animal endomembrane H+-ATPase. This cDNA encodes a hydrophilic polypeptide containing a putative ATP binding site. Lack of a secretion signal sequence suggests it is not processed through the endoplasmic reticulum but translated on cytosolic ribosomes. Comparison of protein sequences shows the 57-kDa subunit from Arabidopsis to be nearly identical with the corresponding subunit in Neurospora vacuolar membrane H+-ATPase, very similar to the beta subunit of the archaebacterium Sulfolobus, and slightly, but nevertheless significantly, homologous to the alpha and beta subunits of the F0F1-ATPases. These results suggest that these different classes of ATPases have evolved from a common ancestor.