Most reservoirs consist of natural and artificial fractures, including isolated microscopic fissures. These fractures form complicated paths for reservoir characterization and fluid movement that ultimately impacts production performance and ultimate recovery. That's why recovery estimations of fractured reservoirs are considered to be extremely challenging due to complexity and heterogeneity of the geological patterns.Oil production from fractured reservoirs results in varying saturation values throughout the reservoir. This is due to the microscopic fissures and heterogeneity of the fracture environment, which could not be swept thoroughly. Higher production/injection ratios also enhance the fingering effect by passing oil through the reservoir.In this study, naturally and artificially fractured cores were used. Analytical and experimental calculations were performed in order to understand the physical structure of the cores. Moreover, in the literature the analytical solutions of the fractures were defined by cubic law (CL), local cubic law (LCL) and modified cubic law (MCL). However, due to the lack of consideration of fracture properties with limited experimental work, there needs still work on this field.Equivalent fracture aperture (EFA) measurements were done by microscope and compared with the analytical calculations. Laboratory results were defined by the equation developed for predicting equivalent fracture apertures as Improved Cubic Law (ICL). Shrinkage in equivalent fracture aperture was also defined by ICL observed by microscope.Using ICL, analytical calculations were done in different environments. Equivalent fracture apertures were calculated for all the experimental flow rates under laminar flow. ICL has worked with the fractured cores flow definition whereas not for homogeneous cores.That's why due to higher calculated values and the 15% error on EFA, a correction coefficient (C) included to the CL equation for the tortuosity or the roughness effect. This result was in th error range of previous results and gave more realistic estimation with respect to the previous studies carried out. Finally, C coefficient was calculated in a range of 0.53–0.65.