In unconventional, naturally fractured reservoirs, networks of structural discontinuities largely control fluid flow. In this study, we mapped and analysed systematic fracture patterns within the Mesozoic succession of Central Spitsbergen to characterise the reservoir-caprock system explored for geological CO2 storage by the Longyearbyen CO2 Lab project. We carried out and integrated structural and stratigraphic analyses of outcrop and borehole data, subdividing the investigated sedimentary interval into five litho-structural units (LSUs): (A) massive to laminated shales characterised by predominant low-angle fractures, (B) heterogeneous, fine-grained intervals with both low- and high-angled fractures, (C) massive, coarsegrained intervals dominated by high-angle fractures, (D) igneous intrusions characterised by syn- and post-emplacement fractures and veins, and (E) carbonate beds dominated by high-angle fractures and veins. LSUs are identified on the basis of their fracture associations, lithologies and dominant sedimentary facies, and thus implicitly include information on the primary porosity and permeability. In general, two main, subvertical extensional fracture sets are recognised: (i) a principal fracture set trending approximately NE–SW to ENE–WSW (J1) and (ii) a subordinate fracture set trending about NNW–SSE to NNE–SSW (J2). Conjugate shear fractures (S1) are trending roughly NE–SW and NW–SE in the coarser-grained and more cemented lithologies. A low-angle fracture set (S2) striking approximately NNW–SSE to WNW–ESE is also observed. Variations in fracture patterns suggest that the LSUs are pseudo-mechanical units, which are able to steer, baffle or impede horizontal and vertical fluid migration due to their primary matrix (i.e., grain size and mineralogy) and fracture network properties. At a larger scale, the resultant stratigraphic and structural architecture controls the hydrogeological regime of the investigated reservoir-caprock succession, providing: (1) fracture-related secondary porosity and permeability, (2) enhanced microfracturing matrix connectivity, and (3) preferential directions of subsurface fluid-flow pathways. We conclude that, given the present-day stress field, subsurface fluid flow would be augmented in an ENE–WSW trend, with possible additional NE–SW communication.