Charged hadrons (protons and heavier ions) have very definite advantages over photons as far as radiotherapy applications are concerned. They allow for a much better spatial dose localization due to their charge, relatively high mass and nature of the energy deposition process. In the frame of an attempt to promote the introduction of hadrontherapy in Argentina we have installed and started using an external beam facility at our tandem accelerator TANDAR. The advantages of heavy ions can only be fully exploited for tumors of well defined localization. In certain types of malignancies, however, the region infiltrated by tumor cells is diffuse, with no sharp boundaries and with microscopic ramifications. In such cases (particularly in certain brain cancers) a more sophisticated scheme has been suggested called boron neutron capture therapy (BNCT). In this work, the use of the Tandar accelerator to produce neutrons for feasibility studies for BNCT through low-energy proton beams on a thick LiF target is being briefly described. Studies on the ¹³C(d,n) reaction and a comparison with other neutron-producing reactions are also mentioned. Simulation work to optimize an accelerator-based neutron production target is discussed. A project is being prepared to develop a small proton accelerator in Argentina. Technical specifications of this machine will be briefly discussed.