A method for the prediction of the response of a wing to nonstationary buffet loads is presented. The time history of the applied load is segmented into a number of time intervals. In each time segment, the nonstationary load is represented by the product of a deterministic shaping function and a statistically stationary random function. An approximate modelling of the load on the wing is given. The wing is divided into panels or elements, and the load is computed from measured or estimated pressure fluctuations at the centre of each panel. A series representation, with terms of the correlated noise type, is used to curve fit the experimentally determined complex buffet pressure power spectral densities. Using the correlated noise form of power spectral density for the random part of the applied load, analytic expressions are derived for the mean square displacement and acceleration response of the wing. An illustration using data available for the F-4E aircraft is included.