[en] The particle beam parameters needed for inertial fusion can be achieved with conventional accelerator technology if heavy ion machines attain the level of performance of the most intense high energy proton machines. Many of the problems posed by this goal pertain to the low energy portions of the accelerator system. In particular, the implied particle current in the rf linac is 10³--10⁴ times the values achieved with existing heavy ion machines. Much of this discrepancy is simply attributable to the great differences between the design considerations relevant to accelerators for fusion and those which have determined the performance of the existing machines. The basic concept chosen at Argonne National Laboratory is cavities containing single drift tubes mounted on lambda/4 supports. Such structures pose the least problem for the beam transport system, and one cavity is placed between adjacent quadrupole magnets. The average voltage gain of the first cells of the low velocity section is moderate; and, although probably acceptable and improved by the end of the 10 MV section, the low initial gain adds to the motivation provided by the transport problem to increase the preinjector voltage substantially above 750 kV