Within the EURADOS (European Radiation Dosimetry Group) self-sustained programme of regular intercomparisons (ICs) for individual monitoring, three campaigns had been performed in 2012, 2017 and 2022 to assess the performance of neutron personal dosemeters routinely used to measure personal dose equivalent, H_p(10). The irradiation schemes were established to provide participants with useful information regarding linearity, reproducibility and response of their dosimetry systems at different energies and angles of incidence. Small alterations were made in irradiation plans for subsequent exercises. Combinations of standard calibration fields and simulated workplace fields were employed, using bare and moderated neutron sources as well as monoenergetic and thermal neutron fields. Neutron energies used to evaluate dosemeter performance ranged from thermal to several MeV. Dose values varied between 0.3 mSv and 15 mSv. The number of participating laboratories in the three campaigns was stable: 31 in IC2012n, 32 in IC2017n and 29 in IC2022n. The dosimetry systems could have been categorized according to their detection principle into albedo and track dosemeters, or a combination of both. In 2012, a category ‘other’ had to be introduced to consider active personal dosemeters and fission track detectors. In 2012, participants were asked to provided results in a two-step procedure: (i) with no information on the radiation fields, and (ii) after receiving additional simplified a priori information on the energy distribution of the neutron fields to enable correction for dosemeter response. In 2017 and 2022, based on a one-step procedure, however, participants might have requested a priori information upon registration, which was reported in the certificate of participation. Intercomparison results were analysed against the acceptance criteria of the ISO 14146:2018 standard, although there were no standardized performance requirements for neutron personal dosemeters at the time of IC2012n. The paper compares the results of the three exercises, identifies major problems for individual monitoring services and shows trends in dosimetric performance. In all three intercomparisons, a few systems had shown problems related to calibration. Data do not allow to draw a universally applicable conclusion with regard to potentially superior performance of one particular dosimetry system.