Appropriate scaling methods for wind tunnel modelling of building internal pressures induced through a dominant opening were investigated. In particular, model cavity volume distortion and geometric scaling of the opening details were studied. It was found that while model volume distortion may be used to scale down buildings for wind tunnel studies on internal pressure, the implementation of the added volume must be done with care so as not to create two cavity resonance systems. Incorrect scaling of opening details was also found to generate incorrect internal pressure characteristics. Furthermore, the effective air slug or jet was found to be longer when the opening was near a floor or sidewall as evidenced by somewhat lower Helmholtz frequencies. It is also shown that tangential flow excitation of Helmholtz resonance for off-centre openings in normal flow is also possible.