MscL, a 15 kDa transmembrane protein, is the only component involved in the formation of a 3 nS channel in the inner membrane of Escherichia coli that opens in response to mechanical or osmotic stress. While previous data had suggested that the functional MscL complex might be a hexamer, a recent crystallographic study of the MscL homologue from M. tuberculosis reveals a pentameric structure. The present work further examines the stoichiometry of the E. coli MscL using a variety of biochemical approaches. Detergent-purified 6His-MscL in solution and MscL in the membrane could be chemically crosslinked with the products displaying ladderlike patterns on SDS gels. Three crosslinking agents (EDC, DMS, and DMA) used at saturating concentrations invariably generated pentamers as the largest product. DSS produced additional bands corresponding to larger complexes although the pentamer band appeared to be the predominant product at high levels of crosslinker. It is not clear whether these extra bands reflect a difference in the crosslinking chemistry of DSS or whether its spacer arm is the longest of those used, or a combination of both facts. For the detergent-solubilized 6His-MscL both sedimentation equilibrium and gel chromatography showed the presence of multiple species. Thus the longer spacer arm could permit both intra- and intercomplex linkages. Nonetheless, the patterns obtained with all agents are consistent with and strongly suggest a pentameric organization for the MscL channel. Expression of MscL as genetically engineered double or triple subunit tandems yields low numbers of functional channels as compared to expressed monomers. The double-tandem assemblies must have an even number of subunits and crosslinking in the membrane confirmed hexamerization. Gel chromatography clearly demonstrated that the channels formed from the double tandems were larger than those formed from WT MscL, consistent with the native channel being pentameric. The observation that both double and triple tandems form channels of normal conductance implies that the pentameric assembly is to some degree independent of the number of subunit repeats in the polypeptide precursor. The channel is thus a pentameric core with the 'extra' subunits left out of the functional complex. From sedimentation equilibrium and size-exclusion chromatography, we also conclude that MscL complexes are not in a dynamic equilibrium with monomers, but are pre-assembled; and thus, their gating properties must result from changes in the conformation of the entire complex induced by the mechanical stress.