Two new forms of the plasma membrane ATP-ase of Micrococcus lysodeikticus NCTC 2665 were isolated from a sub-strain of the microorganism by polyacrylamide gel electrophoresis. One of them had a mol.wt of 368,000 and a very low specific activity (0.80 mumol.min-1.mg protein-1) that could not be stimulated by trypsin. This form has been called B1 (strain B, inactive). If the elctrophoresis was carried out in the presence of reducing agents (i.e., dithiothreitol) and the pH of the effluent maintained at a value of 8.5 another form of the enzyme was obtained. This had a mol.wt of 385,000 and a specific activity of 2.5-5.0 mumol.min-1.mg protein-1 that could be stimulated by trypsin to 5-10 mumol.min-1.mg protein-1. This preparation of the ATPase has been called from BA (strain B, enzyme active). The subunit composition of both forms has been studied by sodium dodecyl sulphate and urea gel electrophoresis and compared to that of the enzyme previously purified from the original strain (form A). The three forms of the enzyme had similar beta and delta subunits, with mol.wt of about 50,000 and 30,000 dalton, respectively. They also had in common the component(s) of relative mobility 1.0, whose status as true subunit(s) of the enzyme remains yet to be established. However, subunit alpha, that had a mol.wt of about a 52,500 in form A (ANDREU et al. Eur. J. Biochem. (1973) 37, 505-515), had a mol.wt similar to beta in form B1 and about 60,000 in form BA. Furthermore BA usually showed two types of this subunit (alpha' and alpha") and an additional peptide chain E) with a mol.wt of about 25,000 dalton. This latter subunit seemed to account for the stimulation by trypsin of form BA. Forms BA could be converted to B1 by storage and freezing and thawing. Conventional protease activity could not be detected in any of the purified ATPase forms and addition of protease inhibitors to form BA failed to prevent its conversion to form B1. The low activity form (B1) was more stable than the active forms of the enzyme and also differeed in its circular dichroism. These results show that M. lysodeikticus ATPase can be isolated in several forms. Although these variations may be artifacts caused by the purification procedures, they provide model systems for understanding the structural and functional relationships of the enzyme and for drawing some speculations about its function in vivo.