Recent developments in the wind loading codes for the UK, Australia and Europe have introduced new, more onerous, pressure coefficient data for the design of free-standing walls. These data, derived from wind-tunnel studies conducted in the mid-1980s in the UK and Australia, have been called into question by various interested parties. In 1993, a research programme was initiated to undertake an independent, full-scale study of the wind pressures on free-standing walls in order to critically appraise the new data and to determine reliable design data. The full-scale, variable-geometry experimental facility which includes automatic, rapid, data-logging instrumentation is described. To supplement this full-scale work, CFD investigations in 2 and 3 dimensions have been undertaken at the University of Auckland using the PHOENICS finite volume code, Version 2.1, with a k-epsilon turbulence model. Comparisons are presented which reveal that despite the simplicity of its structural form, the free-standing wall exhibits surprising aerodynamic effects which render it an excellent and highly challenging test case to model computationally.