Because cracks control the global mechanical and transport properties of crystalline rocks, it is of a crucial importance to suitably determine their aperture distribution, which evolves through alteration processes and rock weathering. Due to the high variability of crack networks in rocks, a multiscale approach is needed. The 14 C-PMMA (polymethylmethacrylate) method was developed to determine crack apertures using a set of artificial crack samples with different controlled apertures and tilt angles and also using Monte Carlo simulations. The experiments and simulations show the same result: the estimation of apparent aperture w A was successful regardless of tilt angle, even if the estimates are less accurate for low tilt angles (<30°). The uncertainties on the estimation of the real crack aperture w R arise from the unknown tilt angle. The ability of the 14 C-PMMA autoradiography method to estimate crack aperture distributions in rock samples was successfully confirmed on a sample of Grimsel granodiorite.