Published measurements of the diffusion of hydrogen in fcc metals are summarized with particular emphasis on the isotope effect. The observations are generally consistent with the usual rate theory model but with the energy levels in both the equilibrium site and the saddle point quantized according to an Einstein oscillator. Not all data fit this scheme, however, diffusion in Pd being a notable exception. Published work for the bcc metals shows much less consistency; surface effects and internal trapping at defects and impurities are apparently of over‐riding importance. A summary of measurements of diffusion of hydrogen in liquid metals is also made.