AbstractSteady state operation of fusion reactors avoids many of the cyclic loads and the associated fatigue of pulsed devices. Extensive experimental and theoretical investigations have led to a good understanding of various mechanisms for non-inductive current drive, but underline the inherent difficulties in achieving such an objective. Whilst measured current drive efficiencies are in good agreement with theoretical models, this efficiency is too low to allow economic steady state operation unless very high bootstrap fractions can be realised. Design optimisation for steady state operation of conventional tokamaks leads to higher aspect ratio and qcyl than for a device optimised for high fusion performance, with a knock-on impact on size and cost. The use of radiofrequency and microwaves for heating and current drive has been extensively investigated. Each frequency range has specific characteristics which may be used to the best advantage in localised regions of the plasma; adequate control of the current profile is likely to require more than one system. Each has critical technologies requiring further development, notably, the couplers for ICRF and LHCD, and the sources for ECRH. The first may limit the future application of these frequency ranges in reactor scale devices at the coupled power densities presently foreseen. The successful application of diamond windows is anticipated to lead to rapid progress in the ECRH range of frequencies.