The inner part of a fusion reactor will have to operate at very high neutron loads. In steady-state reactors the minimum fluence before the scheduled replacement of the reactor core should be at least l0-15 Mw.yr/m2. A more frequent replacement of the core is hardly compatible with economic constraints. A most recent summary of the discussions of these issues is presented in Ref. [l]. If and when times come to build a commercial fusion reactor, the availability of information on the behavior of materials and components at such fluences will become mandatory for making a final decision. This makes it necessary an early development and construction of a neutron source for fusion material and component testing. In this paper, we present information on one very attractive concept of such a source: a source based on a so called Gas Dynamic Trap. This neutron source was proposed in the mid 1980s (Ref. [2]; see also a survey [3] with discussion of the early stage of the project). Since then, gradual accumulation of the relevant experimental information on a modest-scale experimental facility GDT at Novosibirsk, together with a continuing design activity, have made initial theoretical considerations much more credible. We believe thatmore » such a source can be built within 4 or 5 years. Of course, one should remember that there is a chance for developing steady-state reactors with a liquid (and therefore continuously renewable) first wall [4], which would also serve as a tritium breeder. In this case, the need in the neutron testing will become less pressing. However, it is not clear yet that the concept of the flowing wall will be compatible with all types of steady-state reactors. It seems therefore prudent to be prepared to the need of a quick construction of a neutron source. It should also be mentioned that there exist projects of the accelerator-based neutron sources (e.g., [5]). However, they generally have two major disadvantages: a wrong neutron spectrum, with a considerable excess of high-energy neutrons, and smaller test volume. In addition their development requires considerable investments into non-fusion-related technologies, whereas the work on plasma-type sources would certainly boost technology of fusion energy. Broad discussion of these issues can be found in Refs. [3, 6, 7].« less