There are three types of future neutrino facilities currently under study, one based on decays of stored beta-unstable ion beams (?Beta Beams?), one based on decays of stored muon beams (?Neutrino Factory?), and one based on the decays of an intense pion beam (?Superbeam?). In this paper we discuss the challenges each design team must face and the R&D being carried out to turn those challenges into technical opportunities. A new program, the Muon Accelerator Program, has begun in the U.S. to carry out the R&D for muon-based facilities, including both the Neutrino Factory and, as its ultimate goal, a Muon Collider. The goals of this program will be briefly described.

Previous RF experiments with normal-conducting cavities have demonstrated that there is a significant degradation in maximum gradient when the cavity is subjected to a strong axial magnetic field. We have developed a model suggesting that a cavity with beryllium walls may perform better than copper cavities. In this paper we outline the issues that led us to propose fabricating a Be-wall cavity. We also discuss a concept for fabricating such a cavity and mention some of the manufacturing issues we expect to face.