<p>In this study petrophysical characteristics of three consecutive sandstone layers of the Lower Cretaceous Hatira Formation from northern Israel were comprehensively investigated and analysed. The methods used were: experimental petrographic and petrophysical methods, 3D micro-CT imaging and pore-scale single-phase flow modelling, conducted in parallel. All three studied sandstone layers show features indicative of high textural and mineralogical maturity in agreement with those reported from the Kurnub Group in other localities in the Levant. The occurrence of cross-bedding in layers enriched in silt and clay, between the quartz arenite rich beds, may suggest a deposition in a fluvial environment. A higher degree of Fe-ox cementation was observed in the top layer contrasting with a low extent of Fe-ox cementation in the bottom layer. Both quartz-arenite layers are located above and below the intermediate 20 cm thick least permeable quartz wacke sandstone layer. The latter presumably prevented the supply of the iron-rich meteoric water to the bottom layer. Evaluated micro-scale geometrical rocks properties (pore size distribution, pore throat size, characteristic (pore-throat) length, pore throat length of maximal conductance, specific surface area, grain roughness) and macro-scale petrophysical properties (porosity and tortuosity) predetermined the permeability of the studied layers. Large-scale laboratory porosity and permeability measurements show low variability in the quartz arenite (top and bottom) layers, and high variability in the quartz wacke (intermediate) layer. These degrees of variability are confirmed also by anisotropy and homogeneity analyses conducted in the μCT-imaged geometry. Qualitative evaluation of anisotropy (based on statistical distribution of pore space) and connectivity (using Euler Characteristic) were correlated with mineralogy and grain surface characteristics, clay matrix and preferential location of cementation. Two scales of porosity variations were found with variogram analysis of the upper quartz arenite layer: fluctuations at 300 μm scale due to pores size variability, and at 2 mm scale due to the appearance of high and low porosity occlusion by ferruginous bands showing iron oxide cementation. We suggest that this cementation is a result of iron solutes transported by infiltrating water through preferential permeable paths in zones having large grains and pores. Fe-ox precipitated as a result of reaction with oxygen in a partly-saturating realm at the large surface area localities adjacent to the preferential conducting paths. The core part of the study is the investigation of macroscopic permeability, upscaled from pore-scale velocity field, simulated by free-flow in real μCT-scanned geometry on mm-scale sample. The results show an agreement with lab petrophysical estimates on cm-scale sample for the top and bottom layers. Estimated permeability anisotropy correlates with the presence of beddings with 2 mm scale variability in the top layer. The results show that this kind of anisotropy rather than a variability at the pore-scale controls the macroscopic rock permeability. Therefore, we suggest that in order to upscale reliably to the lab permeability, a sufficiently large modelling domain is required to capture the textural features that appear at a scale larger than the pore scale. We also discuss imaging and modelling practices able to preserve the characteristics of the pore network during the entire computational workflow procedure, applicable to studies in the fields of hydrology, petroleum geology, or sedimentary ore deposits.</p>