Crosswind Airborne Wind Energy Systems are currently able to reach altitudes of several hundred meters above ground level. Although this is higher than conventional wind turbines, the optimal altitude is limited by the increasing aerodynamic drag of the tether. Simple models for steady-state crosswind flight suggest that for typical wind shear profiles the power loss due to sweeping a longer tether through the air outweighs the power gained by accessing more powerful winds at higher altitude. A possible solution to this problem is represented by the so called “dancing kitež concept, where a single long cable is attached to two shorter cables, each connected to a kite. The kites are flown in such a way that the long cable is kept in a fixed position with respect to the ground, thus not dissipating power by drag forces, and only the two short cables followthe crosswindmotion of the kites. This concept might be the first that is able to reach altitudes of several thousand meters, thus reaching the extreme power densities of the jet streams, allowing to build low cost and powerful wind turbines. Envisioning a take-off system for such a concept is particularly challenging, and having repeatable and robust take-off and landing capabilities is crucial for the success of the dancing kite principle. Extending the dancing kite principle to a rigid wing setup can have several advantages, above all, it allows for simple take-off and landing sequences. For example, attaching the tethers to the wing tip of several drones results in a multi-drone system that can take-off and land on an axisymmetric circular runway. For this purpose, the axisymmetric round-the-pole flight of a single wind drone has been investigated by means of a dynamic model and an experimental test setup. In this work, a simple 3-degreeof- freedom model represents the flight of the wind drone in spherical coordinates. The drone is modelled as a point mass with aerodynamic properties, throttle and pitch control. The model also takes into account the stabilizing effect fromthe lift of the horizontal stabilizer, fromthe pitch angular velocity, and from the restoring pitch moment due to the centre of gravity being below the aerodynamic centre. The experimental campaign demonstrated full autonomous take-off and landing capabilities of a small scale wind drone flying round the pole in an axisymmetric configuration. The passive stability of the flight suggests that autonomous take-off and landing can easily be achieved in a dual drone system.