Computational Fluid Dynamics (CFD) has recently provided the needed improvements in simulation capabilities that allowed enhancing the design of Darrieus wind turbines. While the performance in operating conditions has increased, poor self-starting is still one of the major drawbacks of these machines. In this study, the aerodynamics of Darrieus-turbines during start-up were investigated using a two-dimensional CFD approach. A fluid-structure interaction simulation was carried out using the ANSYS® FLUENT® solver incorporating the sliding mesh technique and enabling the rotational degree of freedom of the Six Degrees of Freedom (6DOF) solver. The issue of translating a fully-resolved flow field into lumped parameters of use to characterize the instantaneous kinematic properties of the airfoils is tackled by means of two velocity sampling methods recently proposed in the literature, i.e. the 2-PointsAverage and LineAverage methods. The results provide a clear estimation of how much the blade local absolute velocity (V∞, L) is dependent on the instantaneous tip speed ratio during the first revolutions of the starting rotor. At λ < 1, the blockage of the slow-moving blades causes sudden acceleration/deceleration in V∞, L. At 1 < λ < 1.5, a significant increase in V∞, L takes place between 90° < θ < 180° where the blade is moving in the direction of the wind. At λ > 1.5, V∞, L decreases significantly during the azimuth positions of 210° < θ < 340°, where it falls in the other blades’ wake zone. It is also confirmed that the drag force significantly contributes to the Darrieus turbine start-up during the initial cycles, up to λ ≤ 1.5; this in turn corroborates the importance of having reliable data not only for lift coefficient of the airfoil but also of drag coefficient, especially at high angles of attack. Overall, the present study is thought to provide a deeper understanding of the aerodynamics of the Darrieus turbine behaviour during starting that will hopefully enable optimization of these machines.