A key concern during stack emissions testing is the sampling port location. The sampling port must be selected so that the velocity measurements truly represent the exhaust gas flow field within the exhaust stack. Usually the port is located near the stack exit. However, in some instances during field testing, concerns have been raised over the location of the port and its influence in accurate measurements. Furthermore, the use of perforated baffles in the silencer stack has also been questioned; it has been suggested that these could cause flow disturbances that are not acceptable. The veracity of the concerns raised can be examined by physically changing the baffle design and the location of the port. However, this approach would not only take an extensive period of time but would also prove expensive. A lower cost approach is to conduct a CFD analysis on the existing silencer stack and then analyze the results to determine the impact of small perturbations on the measurements and the overall flow field. Modifications, if needed, could be made in the solid model of the silencer/stack system. The impact of these modifications could then be tested with the CFD model to determine how the unmodified stack compares to the modified stack and to establish the degree to which the baffle perforations and the port location influence testing results. The paper describes a detailed CFD analysis of the flow pattern in the silencer stack.