This symposium was consequently organized to review the search for coupling procedures. It appears that the situation is quite satisfactory for 2D flows. Very good correlations can be found between the computational results and the experiments provided that the methods include the wake displacement and curvature effects, and adequate treatment of the trailing edge region and of the shock boundary layer interactions. If so, useful and reliable tools are available for engineering purposes. However, the agreement generally deteriorates with the extension of separated regions or when the shock intensity increases and further improvements are still needed, for example, for single or multielement airfoils near maximum lift or at high Mach numbers. In addition it must be stressed that the quality and the accuracy of the experimental data become questionable as the capability of the computational methods to adequately describe more and more complex situations increases. Therefore precise and reliable code validation requires more and more well documented benchmark data. For 3D configurations, the work is much less advanced; only a few studies have been made and a lot of work has still to be done, especially in the transonic regime. However, the first results obtained using the ideas which have proven to be efficient in 2D are encouraging. The problem of separation with formation and rolling-up of vortex sheets which is specific to 3D flows is of primary importance and the studies already made on this subject will have to be pursued and extended in the compressible and transonic regime. Finally, it has been recorded that the need for detailed benchmark experiments is even greater in 3D than in 2D.