Dose perturbation correction factors, (p ), for LiF, CaF₂ and Li₂ B₄ O₇ solid state detectors have been determined using the EGS4 Monte Carlo code. Each detector was simulated in the form of a disc of diameter 3.61 mm and thickness 1 mm irradiated in a clinical kilovoltage photon beam at a depth of 1 cm in a water phantom. The perturbation correction factor (p ) is defined as the deviation of the absorbed dose ratio from the average mass energy absorption coefficient ratio of water to the detector material, ( _en / )_med,det , which is evaluated assuming that the photon fluence spectrum in the medium and in the detector material are identical. We define another mass energy absorption coefficient ratio, ( _en / )_med,det , which is evaluated using the actual photon fluence spectrum in the medium and detector for LiF and CaF₂ rather than assuming they are identical. ( _en / )_med,det predicts the average absorbed dose ratio of the medium to the detector material within 0.3%. When the difference in atomic number between the cavity and the phantom material is large then their photon fluence spectra will differ substantially resulting in a difference between ( _en / )_med,det and ( _en / )_med,det . The value of (p ) calculated using ( _en / )_med,det is up to 27% greater than unity for a cavity of CaF₂ in 50 kV x-rays. When the atomic number of the medium and detector are similar, their photon fluence spectra are similar, and the difference between ( _en / )_med,det and ( _en / )_med,det is small. For instance their difference for LiF is less than 2%. The average mass energy absorption coefficient ratio, ( _en ( )/ )_{w ,LiF} , evaluated using the mean or representative energy, , is up to 8% different from ( _en / )_{w ,LiF} . For calcium fluoride the difference between ( _en / )_{w ,CaF₂} and ( _en ( )/ )_{w ,CaF₂} is up to 42% in the energy range studied.