The variational method in a reformulated Hamiltonian formalism of quantum electrodynamics (QED) is used to derive relativistic wave equations for systems consisting of n fermions and antifermions of various masses. The derived interaction kernels of these equations include one-photon exchange interactions. The equations have the expected Schrödinger non-relativistic limit. Application to some exotic few lepton systems is discussed briefly.Many-body fermion systems with electromagnetic interactions are the substance of atomic physics. Fundamental fermion and antifermion systems with electromagnetic interactions are of particular interest because they are "pure" QED systems, with point-like constituents and no nuclear force or size effects. Examples of such systems include positronium (Ps: e + e − ), muonium (Mu: µ + e − ) and their ions (Ps − : e + e − e − , Mu − : µ + e − e − ) and four-body systems such as Ps 2 (e + e − e + e − ) and Mu 2 (µ + e − µ + e − ). More generally, for "exotic atoms" such as e + PsH, Ps 2 O, Li + Ps 2 and Na + Ps 2 , nuclear size effects are not negligible. These have received attention in the past (e.g., Ref. 2).The problem of describing relativistic bound states in quantum field theory (QFT) was solved many years ago by Bethe and Salpeter (BS), 3,4 at least in principle. However, the BS method is not free of complications, such as the existence of relative-time coordinates, difficulty of implementation for systems of more than two bodies, and in practice, the perturbative treatment of interactions. There are many papers available in the literature that use the BS method, at least for twoand three-body systems. For example, Adkins and co-workers 5,6 have used BS formalism for the calculation of recoil corrections to the energy levels of hydrogenic ions, and a discussion of issues that will have to be treated for the many-electron case, where highly accurate experiments have been carried out, is given. Using a different approach than the BS formalism, Barut 7 summarized his previous work (including work with his co-authors) and generalized his two-body QED equation to many-body particles interacting via the exchange of massless vector bosons. In his formulation, 7 the relativistic many-body problem has a structure that is similar to that of the Schrödinger many-body problem.An alternative to the BS and other approaches is the variational method within the reformulated Hamiltonian formalism of QFT, introduced by Darewych. 8-10 Among the appealing features of this approach is that it is straightforwardly generalizable to systems of more than two particles, and it can be cast in the form of a relativistic generalization of the Schrödinger description of n-body systems. As a variational method it is applicable, at least in principle, to strongly coupled systems for which perturbation theory may be unreliable. It has disadvantages as well, particularly in that it may not be manifestly covariant, and like all variational methods, the construction of realistic yet tractable trial states may be...