The poor agreement in the photon calibrations of tissue-equivalent ionisation chambers between different participants in two major neutron dosimetry intercomparisons is noted. The methods of converting measurements made with a calibrated secondary standard exposure meter to the absorbed dose in the wall of a tissue-equivalent chamber are discussed. Calibrations of a TE ionisation chamber were made in five photon beams (250 kV, 300 kV, 137Cs, 60Co and 4 MV) and the different calibration factors were compared both for in-air and in-phantom measurements. Good agreement was found in these factors provided that allowance was made for the attenuation and scattering in the wall and build-up cap of the TE chamber for the in-air measurements, and that revised values of C lambda were used for the in-phantom measurements at the two higher energies.