The French studies on space nuclear power systems were relatively active the past six years. It was anticipated that 20‐kWe should have to be supplied to a spacecraft as soon as in 2005 and a reference near term liquid metal‐cooled reactor using available terrestrial technologies as much as possible was mainly investigated. A Brayton cycle heat conversion has been adopted from the beginning and it remains considered. Because first applications are delayed, more attractive concepts can be contemplated. The basic idea is to take advantage of the Brayton cycle specific properties and for instance to use them for the temperature conditioning of the moderator of a thermal spectrum reactor. At first, the utilization of ZrH was thought mandatory, but recent proposals have prompted to use the conventional, effective light water material for that purpose. A gas cycle high pressure (HP) derivation technique makes it possible to maintain water within an adequate temperature range. A Brayton cycle adaptation and an example of a gas‐cooled, particle bed fuel elements, H2O moderated reactor are briefly described. Key comparison data are given. Such concepts should be attractive from fuel inventory, mass, radition shielding and control points of view.