In relation to mechanisms involved in functional recovery of manual dexterity from cervical cord injury or from motor cortical injury, our goal was to determine whether the movements that characterize post-lesion functional recovery are comparable to original movement patterns or do monkeys adopt distinct strategies to compensate the deficits depending on the type of lesion? To this aim, data derived from earlier studies, using a skilled finger task (the modified Brinkman board from which pellets are retrieved from vertical or horizontal slots), in spinal cord and motor cortex injured monkeys were analyzed and compared. Twelve adult macaque monkeys were subjected to a hemi-section of the cervical cord (n = 6) or to a unilateral excitotoxic lesion of the hand representation in the primary motor cortex (n = 6). In addition, in each subgroup, one half of monkeys (n = 3) were treated for 30 days with a function blocking antibody against the neurite growth inhibitory protein Nogo-A, while the other half (n = 3) represented control animals. The motor deficits, and the extent and time course of functional recovery were assessed. For some of the parameters investigated (wrist angle for horizontal slots and movement types distribution for vertical slots after cervical injury; movement types distribution for horizontal slots after motor cortex lesion), post-lesion restoration of the original movement patterns ("true" recovery) led to a quantitatively better functional recovery. In the motor cortex lesion groups, pharmacological reversible inactivation experiments showed that the peri-lesion territory of the primary motor cortex or re-arranged, spared domain of the lesion zone, played a major role in the functional recovery, together with the ipsilesional intact premotor cortex.