Transcranial magnetic stimulation (TMS) studies report that viewing a given action performed by a model activates the neural representation of the onlooker's muscles that are activated during the actual execution of the observed action. Here we sought to determine whether this mirror observation-execution facilitation reflects only muscular specificity or whether it is also influenced by postural congruency between onlooker/model body parts. We recorded motor potentials evoked by single-pulse TMS from the first dorsal interosseous (FDI) and abductor digiti minimi (ADM) muscles during observation of the right index and little finger abduction/adduction movements of models who kept their hands in a palm-down or palm-up position. Moreover, in different experiments observers kept their right hand palm down or palm up. Selective motor facilitation was observed during observation of movements that map the motor function of the targeted muscles, regardless of the posture of the observed hand. Modulation of FDI, however, was obtained only when participants kept their hand palm down; by contrast, modulation of ADM was obtained only when participants kept their hand palm up. Interestingly, electromyographic recordings showed that FDI is mostly active when index abduction/adduction movements are performed in the palm-down position, whereas ADM is mostly active when little finger abduction/adduction movements are performed in the palm-up position. Results show that the influence of the onlooker's hand posture is comparable in action execution and observation, thus indicating a fine-grain functional correspondence between these two processes.