The present paper summarizes the experience accumulated at the Laboratory for Reactor Physics and Thermal-Hydraulics (LRT) of the Paul Scherrer Institute in the field of reactor dosimetry, as well as outlines the recent progress achieved in relation to the associated nuclear data uncertainty propagation methodologies. For dosimetry simulations, the CASMO/SIMULATE/MCNP/FISPACT-II system of codes is in operation at LRT/Nuclear Energy and Safety (NES) Research Division, which is based on the use of validated CASMO/SIMULATE cycle-specific core-follow models of Swiss LWRs. Coupling with FISPACT-II provides the capability for detailed isotopic inventory tracking under irradiation, assessment of materials activation and dpa values, etc. The use of the seamless calculation scheme with translation of the core-follow simulation results into the detailed neutron source specifications for consequent Monte Carlo simulations, accomplished with nuclear data uncertainty propagation capabilities and integrated with appropriate dosimetry validation database, makes the PSI methodology well aligned with the generic best estimate plus uncertainty (BEPU) approach principles. For specific illustrations, this paper presents some results on evaluation of: 1) a few well-known OECD/NEA reactor shielding experimental benchmarks (SINBAD benchmarks H.B. Robinson-2, ASPIS-PCA REPLICA), 2) some dosimetry data obtained from a Swiss PWR and 3) simulation of the dosimetry measurements for the 'PETALE' experimental program at the EPFL research reactor 'CROCUS', as they were foreseen at the time of the experimental planning. Finally, ways to further enhance the simulation methodology and models are discussed.