High energy electron linear accelerators producing photon beams with energies higher than 10 MeV are widely used in radiation therapy. In these beams, fast neutrons are generated, which results in undesired contamination of the therapeutic beam. All modern radiotherapy modalities aim to be highly conformal, which is achieved by using many small fields and longer beam-on times. The beam-on time is proportional to the additional dose from photoneutrons. Hence, it is important to determine the full radiation field correctly in order to evaluate the exposure of patients and medical personnel. For that purpose, in this study solid state nuclear track detectors and Monte Carlo simulations were used. In this study a special emphasis is given to small vaults which were originally built for cobalt machines. After decommissioning of cobalt machines linear accelerators were placed in these vaults. Because of the difference in photon energies between the devices, shielding of the vault had to be enhanced. During the reconstruction of 60^Co vaults for linear accelerators, the main limitation was the space. The vault walls had to be enhanced for photon shielding, but there was no space for adding more concrete to the walls. Therefore, lead and steel panels were added into the walls instead. These panels can be sources of photoneutrons. One of the goals in this study was to decrease measurement uncertainty when using solid state nuclear detectors by automatization of track counting. Track counting was automatized by using software ImageJ. Measurement uncertainty was not decreased, but the process of track counting is now less time consuming which is important, especially when there are many detectors to be processed. The goal was to decrease measurement uncertainty caused by detector energy dependence by calibrating detectors in known neutron spectra. Neutron spectra were determined using MCNP code for Monte Carlo simulations. Model of accelerator was built using MCNP code in such way that calculated photon dose profiles and percentage depth doses fit to measured data. Neutron source strength is calculated and is 1.12∙10^12 neutrons per Grey photon dose. Linear accelerator head cover is included in Monte Carlo model and the results show that omitting head cover from model can cause neutron fluence underestimation and neutron energy overestimation. That can lead to large uncertainties in estimation of neutron dose to patients and staff. The vault in which accelerator is placed is also modelled using MCNP code. Neutron spectra in measuring positions inside and outside the vault are calculated. In the same positions detectors were irradiated and calibrated against known neutron spectra. One of the goals was to compare measurements and simulations and investigate if there was a possibility to determine neutron energy spectra directly from the shape and size of tracks on detector. Most of the tracks on detectors are caused by thermal neutrons and their energy is insignificant compared to the much larger energy produced by nuclear reaction. Therefore, it is impossible to know the energy of incoming neutrons and because of that further analysis of track shape and size was not continued. Neutron fluence in patient plane is calculated and the influence of beam size on fluence is investigated. Neutron dose equivalent in isocenter is estimated and it is 3.3 mSv/Gy photon dose. Neutron equivalent dose for staff at operator console was estimated and it is 324 μSv per year. The hypothesis was that contribution to the dose from neutron radiation should be included in radiotherapy planning. The estimation is that neutron equivalent dose for a prostate patient irradiated with 74 Gy photon dose can be 0.3 Sv. That is not negligible and neutron contribution to overall dose should be taken into account. Our further investigation will be focused on estimating neutron dose to organs at risk by using solid state nuclear detectors calibrated against spectrum in patient plane. Another hypothesis was that there is a need for personal neutron dosimetry in radiotherapy, especially in the vicinity of small vaults that were reconstructed after cobalt machine decommissioning. Neutron dose equivalents outside the vault are far from exposure limits and therefore there is no need for neutron personal dosimetry. Kersey and Wu-McGinley empirical equations are the most commonly used for estimation of neutron dose equivalent at the maze door in the vault. In this study Monte Carlo simulations of various geometrical and compositional changes of vault were conducted to assess accuracy of these equations. The empirical equations gave results with reasonable accuracy when vaults were of the usual size. When the vault is decreased to the size usually used for 60^Co unit vault, the most commonly used equations showed significant difference in results in comparison to Monte Carlo simulations. Therefore, for vaults limited in size, a new simplified equation is suggested to assess the neutron dose equivalent at the maze door.