The development of wind tunnel sections with partly open walls has made it possible to operate at continuously variable speeds through the transonic range up to low supersonic speeds. Some consideration must be given, however, to the general reliability of the measurements in these tunnels of the aerodynamic characteristics of air vehicles or missiles. A comparative assessment of the possible overall errors in force and pressure measurements in relation to desirable standards of accuracy is, in fact, required. A number of points connected with such an assessment are briefly discussed (mainly on the basis of experience in the 3 ft. tunnel at R.A.E., Bedford) in this paper, which touches on flow uniformity, accuracy of model manufacture, and interference effects, including a comparison between tunnel and free flight model results; illustrative examples of results for an aircraft model, with some full scale flight comparisons, are discussed briefly and a few comments are made on the problems associated with the achievement of representative flow conditions on sting–mounted models at transonic speeds.