OBJECTIVE To measure the out-of-field mean photon energy and dose imparted by the secondary radiation field generated by 6 MV and 6 MV FFF beams using TLD-300 and TLD-100 dosimeters and to use the technique to quantify the contributions from the different sources that generate out-of-field radiation. METHODS The mean photon energy and the dose were measured using the TLD-300 glow curve properties and the TLD-100 response, respectively. The TLD-300 glow curve shape was energy-calibrated with gamma rays from 99m Tc, 18 F, 137 Cs, and 60 Co sources, and its energy dependence was quantified by a parameter obtained from the curve deconvolution. The TLD-100 signal was calibrated in absorbed dose-to-water inside the primary field. Dosimeters were placed on the linac head, and on the surface and at 4.5 cm depth in PMMA at 1-15 cm lateral distances from a 10 × 10 cm2 field edge at the isocenter plane. Three configurations of dosimeters around the linac were defined to identify and quantify the contributions from the different sources of out-of-field radiation. RESULTS Typical energies of head leakage were about 500 keV for both beams. The mean energy of collimator-scattered radiation was equal to or larger than 1250 keV and, for phantom-scattered radiation, mean photon energies were 400 keV for the 6 MV and 300 keV for the 6 MV FFF beam. Relative uncertainties to determine mean photon energy were better than 15% for energies below 700 keV, and 40% above 1000 keV. The technique lost its sensitivity to the incident photon energy above 1250 keV. On the phantom surface and at 1-15 cm from the field edge, 80%-90% of out-of-field dose came from scattering in the secondary collimator. At 4.5 cm deep in the phantom and 1-5 cm from the field edge, 50%-60% of the out-of-field dose originated in the phantom. At the points of measurement, the head leakage imparted less than 0.1% of the dose at the isocenter. The 6 MV FFF beam imparted 8-36% less out-of-field dose than the 6 MV beam. These energy results are consistent with general Monte Carlo simulation predictions and show excellent agreement with simulations for a similar linac. The measured out-of-field doses showed good agreement with independent evaluations. CONCLUSIONS The out-of-field mean photon energy and dose imparted by the secondary radiation field were quantified by the applied TLD-300/TLD-100 method. The main sources of out-of-field dose were identified and quantified using three configurations of dosimeters around the linac. This technique could be of value to validate Monte Carlo simulations where the linac head design, configuration, or material composition are unavailable.