This chapter reviews recent advances in the identification of the structural elements of the permeability transition pore. The discovery that cyclosporin A (CsA) inhibits the pore proved instrumental. Various approaches indicate that CsA blocks the pore by binding to cyclophilin (CyP)-D. In particular, covalent labelling of CyP-D in situ by a photoactive CsA derivative has shown that pore ligands have the same effects on the degree to which CsA both blocks the pore and binds to CyP-D. The recognition that CyP-D is a key component has enabled the other constituents to be resolved. Use of a CyP-D fusion protein as affinity matrix has revealed that CyP-D binds very strongly to 1:1 complexes of the voltage-dependent anion channel (from the outer membrane) and adenine nucleotide translocase (inner membrane). Our current model envisages that the pore arises as a complex between these three components at contact sites between the mitochondrial inner and outer membranes. This is in line with recent reconstitutions of pore activity from protein fractions containing these proteins. The strength of interaction between these proteins suggests that it may be a permanent feature rather than assembled only under pathological conditions. Calcium, the key activator of the pore, does not appear to affect pore assembly; rather, an allosteric action allowing pore flicker into an open state is indicated. CsA inhibits pore flicker and lowers the binding affinity for calcium. Whether adenine nucleotide translocase or the voltage-dependent anion channel (via inner membrane insertion) provides the inner membrane pore has not been settled, and data relevant to this issue are also documented.