When people witness or imagine the pain of another person, their nervous system may react as if they were feeling that pain themselves. Early neuroscientific evidence indicates that the firsthand and vicarious experiences of pain share largely overlapping neural structures, which typically correspond to the lateral and medial brain regions that encode the sensory and the affective qualities of pain. Such neural circuitry is highly malleable and allows people to flexibly adjust the empathic behavior depending on social and personal factors. Recent views posit, however, that the brain can be conceptualized as a complex system, in which behavior emerges from the interaction between functionally connected brain regions, organized into large-scale networks. Beyond the classical modular view of the brain, here we suggest that empathic behavior may be understood through a dynamic network-based approach where the cortical circuits associated with the experience of pain flexibly change in order to code self- and other-related emotions and to intrinsically map our mentality to empathetically react to others.