Although theoretical discourse and experimental studies on the self- and reward-biases have a long tradition, currently we have only a limited understanding of how the biases are represented in the brain and, more importantly, how they relate to each other. We used multi-voxel pattern analysis to test for common representations of self and reward in perceptual matching in healthy human subjects. Voxels across an anterior–posterior axis in ventromedial prefrontal cortex (vmPFC) distinguished (i) self–others and (ii) high–low reward, but cross-generalization between these dimensions decreased from anterior to posterior vmPFC. The vmPFC is characterized by a shift from a common currency for value to independent, distributed representations of self and reward across an anterior–posterior axis. This shift reflected changes in functional connectivity between the posterior part of the vmPFC and the frontal pole when processing self-associated stimuli, and the middle frontal gyrus when processing stimuli associated with high reward. The changes in functional connectivity were correlated with behavioral biases, respectively, to the self and reward. The distinct representations of self and reward in the posterior vmPFC are associated with self- and reward-biases in behavior.