Humans usually point at objects to communicate with other persons, although they generally avoid pointing at the other's body. Moreover, patients with heterotopagnosia after left parietal damage cannot point at another person's body parts, although they can point at objects and at their own body parts and although they can grasp the others' body parts. Strikingly, their performance gradually improves for figurative human body targets. Altogether, this suggests that the body of another real person holds a specific status in communicative pointing. Here, we test in healthy individuals whether performance for communicative pointing is influenced by the communicative capacity of the target. In Experiment 1, pointing at another real person's body parts was compared to pointing at objects, and in Experiment 2, the person was replaced by a manikin. While reaction times for pointing at objects were shorter compared to pointing at other person's body parts, they were similar for objects and manikin body parts. By adapting Experiment 1 to PET-scan imaging (Experiment 3), we showed that, compared to pointing at objects, the brain network for pointing at other person's body parts involves the left posterior intraparietal sulcus, lesion of which could cause heterotopagnosia. Taken together, our results indicate that the specificity of pointing at another person's body goes beyond the visuo-spatial features of the human body and might rather rely on its communicative capacity.