While there have been several studies investigating the neural correlates of action observation associated with hand grasping movements, comparatively little is known about the neural bases of observation of reaching movements. In two experiments, using functional magnetic resonance imaging (fMRI), we defined the cortical areas encoding reaching movements and assessed their sensitivity to biological motion and to movement velocity. In the first experiment, participants observed video-clips showing either a biological effector (an arm) or a non-biological object (rolling cylinder) reaching toward a target with a biological and a non-biological motion, respectively. In the second experiment, participants observed video-clips showing either a biological effector (an arm) or a non-biological object (an arrow) reaching toward a target with the same biological motion profiles. The results of the two experiments revealed activation of superior parietal and dorsal premotor sites during observation of the biological motion only, independent of whether it was performed by a biological effector (reaching arm) or a non-biological object (reaching arrow). These areas were not activated when participants observed the non-biological movement (rolling cylinder). To assess the responsiveness of parietal and frontal sites to movement velocity, the fMRI repetition-suppression (RS) technique was used, in which movement was shown with same or different velocities between consecutive videos, and observation of identical stimuli was contrasted with observation of different stimuli. Regions of interest were defined in the parietal and frontal cortices, and their response to stimulus repetition was analyzed (same vs. different velocities). The results showed an RS effect for velocity only during the observation of movements performed by the biological effector and not by the non-biological object. These data indicate that dorsal premotor and superior parietal areas represent a neural substrate involved in the encoding of reaching movements and that their responsiveness to movement velocity of a biological effector could be instrumental to the discrimination of movements performed by others.