Purpose: Specialized treatment machines such as the CyberKnife, TomoTherapy, or the GammaKnife, utilize flattening filter free (FFF) photon beams and may not be able to generate a 10 cm x 10 cm reference field. A new Code of Practice has recently been published (IAEA TRS483) to give recommendations for these machines. Additionally, some standard laboratories provide measured beam quality correction factors for the user’s reference chamber, which can be used instead of the published tabulated beam quality correction factors. The purpose of this study was first to assess how different recommendations, as outlined below, affect the reference dosimetry at the CyberKnife, and second, to assess the impact of using measured rather than tabulated beam quality correction factors on measured dose. Methods: Following recommendations in TRS398, three field chambers (IBA CC04, Exradin A19, and Exradin A12S) were cross-calibrated with a user’s reference chamber (IBA FC65-G), which was calibrated in a Cobalt-60 (Co-60) beam by a primary standards laboratory. The chamber response was corrected for influence quantities such as temperature, pressure, ion recombination, polarity, and beam quality. Additionally, correction factors for volume averaging and differences due the FFF beam spectrum were determined for the CyberKnife beam. Three different methods were utilized - TRS398; Intermediate (i.e. TRS398 with additional published recommendations); and TRS483. The measurements were undertaken in a 10 cm × 10 cm field defined by jaws for a uniform flattened (WFF) 6 MV photon beam of a Varian TrueBeam linear accelerator (linac) with a source to detector distance (SDD) of 100 cm, and in a 60 mm diameter circular field for a 6 MV flattening filter free (FFF) Accuray CyberKnife beam with SDD of 80 cm. All measurement was performed at 10 cm deep in a full scatter phantom as defined in TRS398. Results: Differences between the three methods in volume averaging correction factors ranged from 0.01% to 0.45% depending on the chamber assessed. As expected, an increased chamber length leads to a larger correction factor. The differences in beam spectrum correction factors range from 0.09% to 0.3%. Negligible differences in beam quality correction factors were observed; however, differences up to 1% were found between measured and tabulated values. Differences in cross-calibrated chamber calibration coefficients range from 0.05% to 0.51% depending on the chamber assessed. Differences in measured dose are up to 0.87% between Method TRS398 and Intermediate, again chamber dependent, and 0.28% between Method Intermediate and TRS483. Conclusion: Using chambers cross-calibrated in the linac beam can lead to differences in measured dose per Monitor Unit (MU) in the CyberKnife beam of approximately 0.5% between chambers. Using Method Intermediate vs using recommendations given in TRS483 led to a difference of 0.28% in measured dose per MU, which is due to differences in volume averaging and beam spectrum correction factors. Using TRS483 is recommended as the cross-calibration is done in the CyberKnife beam and accounts for its specific reference conditions. It will also ensure consistency between different centers. The measured beam quality correction factors agree within the uncertainties with the tabulated values.