One of the key set of uncertainties in understanding the reservoir for flow simulation purposes is the location, geometry and properties of the internal discontinuities to fluid or pressure communication. Barriers are created by faults, cemented fractures, lithology changes, or boundaries between individual genetic flow units. Description and evaluation of these features are necessary for a precise understanding of the dynamic behaviour of the reservoir, affecting predictions of permeable fairways, bypassed hydrocarbons, and hence production. Furthermore, missing, mis-located, or mis-represented barriers can lead to inadequate predictions of reservoir performance and hence non-optimal reservoir management and hydrocarbon recovery. Compartmentalization and the presence of barriers can, to some extent, be recognized during production by unusually low pressures, early liberation of solution gas, and poor pressure support from injectors, surrounded by high pressures. Additionally, once the simulation model has been constructed, this can be used to position and quantify expected barriers by history matching their properties to the produced volumes and pressures. Contributing to this process is 3D seismic and geological interpretation, which provide information on the likely positions of potentially important structural discontinuities, although these data cannot help in directly constraining dynamic flow performance. In such a process subtle connectivities between sand channels are often lost and local high permeability pathways cannot be accurately quantified. Fault seal capacity can be estimated considering the fault zone heterogeneity via reservoir juxtaposition and shale gouge ratio calibrated by laboratory tests (Manzocchi et al., 1999). A further technique to determine the degree of communication across the barriers for engineering purposes is well-well transient interference testing (Stewart et al., 1984). Reservoir connectivity and the presence of flow boundaries in the reservoir volume can also be assessed by interpreting transient analysis and simulation of extended well tests in combination with the initial pressure data from repeat formation tests (for example, Richardson et al., 1997). Techniques based on well data for pressure information lack aerial coverage and resolution and, in principle, this gap can be filled by 4D seismic which could supply the extra detailed information required for a general reservoir connectivity assessment at sub-simulator cell resolution. Past case studies have highlighted this value. For example, Sonneland, et al. (2000) used a saturation-based approach to assess seal along a fault network with 4D seismic in the Gullfaks field. Barkved et al. (2003) indicated that a sealed fault block, previously unrecognized on the 3D seismic, could be detected after pressure depletion on Valhall. Parr and Marsh (2000) gave examples of the visual location of barriers inferred from 4D signature evolution due to horizontal producers and also the added information on barrier breakdown during production in West of Shetland data. Finally, Almaskeri and MacBeth (2004) showed how discontinuities in the 4D seismic signatures might be used as a fine-scale attribute to reveal the location of barriers. Building on this latter work, the aim here is to provide a further evaluation of barriers by applying a recently developed technique for horizontal permeability estimation (MacBeth and Almaskeri, 2005). Specifically, by modifying the way in which the estimation technique is implemented, it is possible to focus on the low magnitude components of the permeability spectrum related predominantly to the barrier contribution, and then to use the magnitude of the permeability drops to assess the extent to which the barrier can transmit pressure. This method results in a seismic map of barriers and hence an assessment of the degree of communication.