Abstract This paper quantitatively evaluates the hierarchical structure of multiple-porosity systems, including natural micro-fractures, inter-granular, and intra-granular pores, using nuclear magnetic resonance (NMR) measurements. Conventional well logs cannot distinguish between different pore structures. NMR measurements, although counted among the most reliable methods to measure formation porosity, has been conventionally considered as insensitive to the existence of natural fractures. Thus, the impact of natural micro-fractures on NMR measurement has rarely been investigated. We simulated NMR response in porous media using a random-walk algorithm. We randomly distributed and oriented natural micro-fractures in various porous rock matrices. We then quantified the sensitivity of NMR T2 (spin-spin relaxation time) distribution to the presence of natural fractures within three-dimensional (3D) Micro-Computed Tomography (micro-CT) images of carbonate and sandstone rock samples andsynthetic organic shale matrices. Results from synthetic rock samples showed that NMR T2 distribution can be significantly affected by aperture, concentration, and shape (e.g., needle-like or planar shape) of natural fractures, as well as size and connectivity of inter-granular pores. We also quantified the impact of diffusional coupling effect between the micro-fractures and inter-granular pores on NMR T2 distribution. Applications of this research include reliable reservoir characterization in challenging multiple-porosity systems such as naturally fractured carbonates and organic-rich source rocks, whereconventional well logs cannot reliably characterize the complex pore structure andinterpretation techniques for unconventional logs such as NMR are not fully developed.