This study aimed to identify sites and mechanisms of long-term plasticity following lower limb muscle training. Two groups performing either a postural stability maintenance training (SMT) or a ballistic ankle strength training (BST) were compared to a non-training group. The hypothesis was that practicing of a self-initiated voluntary movement would facilitate cortico-spinal projections, while practicing fast automatic adjustments during stabilization of stance would reduce excitatory influence from the primary motor cortex. Training effects were expected to be confined to the practiced task. To test for training specificity, motor evoked potentials (MEP) induced by transcranial magnetic stimulation (TMS) were recorded at rest and during motor tasks that were similar to each training. Intracortical, cortico-spinal, as well as spinal parameters were assessed at rest and during these tasks. The results show high task and training specificity. Training effects were only observable during performance of the trained task. While MEP size was decreased in the SMT group for the trained tasks, MEP recruitment was increased in the BST group in the trained task only. The control group did not show any changes. Background electromyogram levels, M. soleus H-reflex amplitudes and intracortical parameters were unaltered. In summary, it is suggested that the changes of MEP parameters in both training groups, but not in the control group, reflect cortical motor plasticity. While cortico-spinal activation was enhanced in the BST group, SMT may be associated with improved motor control through increased inhibitory trans-cortical effects. Since spinal excitability remained unaltered, changes most likely occur on the supraspinal level.