Ion Bernstein wave heating (IBWH) utilizes the ion Bernstein wave (IBW), a hot plasma wave, to carry the radio frequency (rf). power to heat tokamak reactor core. Earlier wave accessibility studies have shown that this finite-Larm0r-radius (FLR) mode should penetrate into a hot dense reactor plasma core without significant attenuation. Moreover, the IBW's low phase velocity (m / kt,= VTi << Va) greatly reduces the otherwise ,serious wave absorption by the 3.5 MeV fusion ox-particles. In addition, the property of IBW's that k±p i = 1 make:; l(xcalized bulk ion heating possible at the ion cyclotron harmonic layers. Such bulk ion heating can prove useful in optimizing fusion reactivity. In another vein, with proper selection of parameters, IBW's can be made subject to strong localized electron Landau damping near the major ion cyclotron harmonic resonance layers. This property can be useful, for example, for rf current drive in the reactor plasma core,. IBW's can be excited with loop antennas or with a lower-hybrid like waveguide launcher at the plasma edge, the latter structure being one that is especially compatible with reactor application. In either case, the mode at the plasma edge is an electron plasma wave (EPW). Deeper in the plasma, the EPW is mode-transformed into an [BW. Such launching and mode-transformation of IBW's were first demonstrated in experiments in the ACT-1 plasma torus and in particle simulation calculations. These and other aspects of IBW heating physics have been investigated through a number of experiments performed on ACT-1, JIPPTII-U, TNT, PLT and Alcator-C. In these experiments both linear and nonlinear heating processes have been observed, Interestingly, improvement of plasma confinement was also observed in the PLT, Alcator-C, and JIPPTII-U experiments, opening up the possible use of IBW's for the active control of plasma transport. Two theoretical explanations have been proposed: one ba_d on tour-wave-mixing of IBW with low frequency turbulence, the other on the noninear generation of a velocity-shear laver. Both models are consistent with the observed threshold power level of a few hundred kW in the experiments.. Experiments on lower field plasmas on JI-:I'II-M and DII!-D have raised some concern with the IBW wave-launching process. The experiments showed serious impurity release from the walls but little or no core heating, a combination of circumstances strongly suggestive of edge heafingl Possible parasitic channels could include the excitation of short wavelength modes by the Faraday shield's fringing fields, antenna sheath-wave excitation, an axial convective loss " channel, and nonlinear processes such as parametric instability and ponderomotive effects. Suggested remedies include changes in the antenna phasing, the use of low-Z insulators, operating at higher frequencies, positioning the plasma differently with respect to the antenna, eliminating the Faraday shields, and using a waveguide launcher. The recent JIPPTII-U experiment, employing a ()-rcphased antenna array with a ...