In this rapport, the aeroelastic phenomenon, flutter is investigated for the Great Belt East Bridge. The bridge is examined at two different stages during construction: when only 15% of the bridge deck is mounted and when the bridge is fully erected. A numerical approach, CFD, is used for both stages. According to the well-known Scanlan theory, the aerodynamic loads are described by a set of aerodynamic parameters, which can be derived by use of different approaches. These aerodynamic derivatives are found by a two dimensional analysis of the cross section of the bridge, by a forced oscillation test. The derivatives are used in determination of the critical flutter wind velocity for the fully erected bridge. The purpose of this 2d-test is to validate the CFD as a mean of computing critical flutter wind velocities. The resulting critical flutter wind velocity of 71.9 m/s found, is not far from the velocity found in wind tunnel tests, of 70 m/s to 74 m/s. For the investigation of the bridge during erection, two three dimensional analysis are performed. A forced vibration test is performed in a similar way as for the two dimensional case. In this analysis the fluid is allowed to flow around the free end of the bridge, introducing some end effects. The structural behavior of the bridge during construction shows that the lowest vertical eigenfrequency is very close to the lowest torsional eigenfrequency. This implies that the critical flutter wind velocity should be somewhat lower for the bridge during construction. This velocity is found at 47.6 m/s. Wind tunnel tests of this construction stage show a critical flutter wind velocity of 43.3 m/s. The rapport further investigates the possibility of using a fully coupled fluid-structure interaction when determining critical flutter wind velocities. A two degree of freedom system is used to determine the motion of the bridge deck, which is assumed rigid. A first order backward Euler integration scheme is applied in Ansys 11 when prescribing the movements of the system. The mesh of the fluid domain is very coarse, in order to reduce the computation time. The simulations have been run, but with some difficulties with respect to damping. The test show a critical flutter wind velocity of approximately 20 m/s to 25 m/s which is far lower than the wind tunnel tests. The test is, however described in the report, as the method gives a very good visual and physical understanding of the aeroelastic phenomenon.