When performing movements with different spatial trajectories in both upper limbs simultaneously, patterns of interference emerge that can be overcome with practice. Even though studies on the role of augmented feedback in motor learning have been abundant, it still remains to be discovered how overcoming such specific patterns of spatial interference can be optimized by instructional intervention. In the present study, one group acquired a bimanual movement with normal vision, whereas a second group received augmented feedback of the obtained trajectories on a computer screen in real time. Findings revealed that, relative to normal vision, the augmented feedback hampered skill learning and transfer to different environmental conditions. These observations are discussed in view of the benefits and pitfalls of augmented feedback in relation to task context and instructional condition.