Recent letters by Stephen and by Scully and Lee suggest that the radiation emitted from a small Josephson junction mounted in a finite- Q resonant cavity and biased at a particular dc potential difference V can be shifted from the Josephson frequency ω J = 2 eV/ h ̷ by cavity detuning. This particular conclusion is not correct but stems from use of the electrostatic potential V 0 for the electrochemical potential V in the Josephson formula. Cavity detuning can cause V 0 to differ measurably from V, but it does not shift the radiation frequency relative to ω J. This is consistent with very general thermodynamic arguments dependent upon the relatively weak assumption that the superconductors on each side of the junction are internally in thermal equilibrium.