When tracing a template with mirror-reversed vision (or distorted vision), the sensory information arising from the movement does not match the expected sensory consequences. In such situations, participants have to learn a new visuomotor mapping in order to trace the template with an accuracy and speed approaching that observed when tracing with direct vision. There are several suggestions that such visuomotor learning requires lowering the gain of the proprioceptive inputs. Generally, subjects learn this task in a seated condition offering a stable postural platform. Adapting to the new visuomotor relationship in a standing condition could add complexity and even hinder sensorimotor adaptation because balance control and processing of additional information typically interfere with each other. To examine this possibility, older individuals and young adults (on average, 70 and 22 years of age, respectively) were assigned to groups that trained to trace a shape with mirror-reversed vision in a seated or a standing condition for two sessions. For a third session , the seated groups (young and elderly) transferred to the standing condition while the standing groups continued to perform the tracing task while standing. This procedure allowed comparing the tracing performance of all groups (with the same amount of practice) in a standing condition. The standing groups also did a fourth session in a seated condition. Results show that older participants initially exposed to the standing condition were much slower to trace the template than all other groups (including the older group that performed the tracing task while seated). This slowness did not result from a baseline general slowness but from a genuine interference between balance control and the visuomotor conflict resulting from tracing the pattern with mirror-reversed vision. Besides, the Standing-Old participants that transferred to a seated condition in the fourth session immediately improved their tracing by reducing the total displacement covered by the pen to trace the template. Interestingly, the results did not support a transfer-appropriate practice hypothesis which suggests that training in a standing condition (at the third session) should have benefited the performance of those individuals who initially learned to trace the mirror pattern in a standing condition. This has important clinical implications: training at adapting to new AGE (2014) 36:823–837