The airflow rate of a building ventilated by wind is usually predicted by using the wind pressure coefficients obtained for a sealed building and discharge coefficients based on measuring the airflow characteristics through an opening in a sealed chamber (chamber method). This can result in the underestimation of wind driven flow through large openings located on opposite sides of a room. In this paper, the discharge coefficient, based on the chamber method, and the actual condition of cross-ventilation are calculated and compared with each other by means of stream tube analysis. The driving pressure based on wind pressure coefficients obtained from a sealed building are also compared with those based on pressures inside the stream tube of the actual flow field representing a porous rather than sealed building. A building model of dimensions 120 mm (width)×120 mm (height)×180 mm (length) was used for the analyses. The size of openings, expressed as the porosity (opening area divided by façade area), was 11.6 %, 20.7 % and 46.5 %. These models were analyzed by CFD simulation and the stream tubes caught by the opening were determined. From the analysis the errors in discharge coefficient and wind pressure coefficient were identified. Finally, the flow rate based on these discharge coefficients and driving pressures were calculated and compared. It is shown that the effect of the underestimation of the discharge coefficients by the chamber method is significant for all cases of porosity studied in this paper. Moreover, it is shown that the use of wind pressure coefficients is not appropriate for the case of extremely large openings.