Generation of random deviates for relativistic quantum-statistical distributions

We provide an algorithm for generation of momenta (or energies) of relativistic particles according to the relativistic Bose-Einstein or Fermi-Dirac distributions. The algorithm uses rejection method with effectively selected comparison function so that the acceptance rate of the generated values is always better than 0.9. It might find its use in Monte-Carlo generators of particles from reactions in high-energy physics.