The modal parameters of an aircraft structure are currently estimated from ground vibration tests (GVT). These tests are carried out on ground in order to estimate the frequency response functions first and then the modal parameters. Such estimates require one or more shakers to excite the structure together with simultaneous measurements of both the input and the output signals. Recent developments in operational modal analysis allowed such modal identification from the measurements of the output responses only, provided that the unmeasurable excitation is practically a white noise stochastic input in the frequency range of interest and is randomly exciting the different parts of the structure. In this paper the effects of the different test setup on the modal parameter estimates will be presented. Both input-output based experimental modal analysis and operational modal analysis are performed on an Unmanned Aerial Vehicle, the Clarkson University Golden Eagle. The modal parameters estimated from the standard ground vibration test are compared with the ones achieved by using the engine induced vibrations as means of excitation while the UAV is taxiing.