A persistent challenge in reservoir modeling is to assign representative values for vertical permeability in reservoir models. It is a common practice to use kv/kh directly from the core measurements but this is not at the same scale as the model grid. It is also frequent to use one value for entire zones. The approach of using a single value applied to a single zone when the reservoir is heterogeneous is not correct and the same can be said regarding using core plug kv/kh as the grid scale is larger than the core plug therefore resulting in a difference in the statistical support of the two measurements. This paper describes the technique to understand reservoir vertical connectivity by properly defining kv/kh, which later will be used as a matching parameter during the History Match process. Most of the time the data provided by the plugs is not appropriate due to the absence of vertical sampling or to problems of representativeness. Where plug data is available kv/kh ratios are often found to be close to one, a values too high to be realistic when considering the geological heterogeneity at small scale or plug scale. Even when plug measurements show homogenous at small scale (kv/kh ∼1), at grid cell scale the Kv and Kh are expected to be different (<<1). A methodology is proposed in order to give more appropriate ranges of values to the grid cell size, based on the small-scale permeability anisotrophy ratio (kv/kh) into grid cell size. A novel approach was established to get the ranges of kv/kh based on core mini-k measurements and core-plugs data. This innovative workflow also considers a multi-disciplinary approach using Experimental Design methods to better understand and validate the kv/kh ranges. The main learning from this project is the importance of understanding vertical connectivity in a complex reservoir by properly capturing heterogeneity and other major geological features in the integrated 3D dynamic model with the aim to achieve an optimized history match at reservoir level but also more importantly on a well-by-well basis.