Birds appear to use their tails during flight, but until recently the aerodynamic role that tails fulfil was largely unknown. In recent years delta–wing theory, devised to predict the aerodynamics of high–performance aircraft, has been applied to the tails of birds and has been successful in providing a model for the aerodynamics of a bird's tail. This theory now provides the conventional explanation for how birds' tails work. A delta–wing theory (slender–wing theory) has been used, as part of a variable–geometry model to predict how tail and wing shape should vary during flight at different airspeeds. We tested these predictions using barn swallows flying in a wind tunnel. We show that the predictions are not quantitatively well supported. This suggests that a new theory or a modified version of delta–wing theory is needed to adequately explain the way in which morphology varies during flight.