The present study examines the functional and anatomical underpinnings of egocentric and allocentric coding of spatial coordinates. For this purpose, we set up a functional magnet resonance imaging experiment using verbal descriptions of spatial relations either with respect to the listener (egocentric) or without any body-centered relations (allocentric) to induce the two different spatial coding strategies. We aimed to identify and distinguish the neuroanatomical correlates of egocentric and allocentric spatial coding without any possible influences by visual stimulation. Results from sixteen participants show a general involvement of a bilateral fronto-parietal network associated with spatial information processing. Furthermore, the egocentric and allocentric conditions gave rise to activations in primary visual areas in both hemispheres. Moreover, data show separate neural circuits mediating different spatial coding strategies. While egocentric spatial coding mainly recruits the precuneus, allocentric coding of space activates a network comprising the right superior and inferior parietal lobe and the ventrolateral occipito-temporal cortex bilaterally. Furthermore, bilateral hippocampal involvement was observed during allocentric, but not during egocentric spatial processing. Our results demonstrate that the processing of egocentric spatial relations is mediated by medial superior-posterior areas, whereas allocentric spatial coding requires an additional involvement of right parietal cortex, the ventral visual stream and the hippocampal formation. These data suggest that a hierarchically organized processing system exists in which the egocentric spatial coding requires only a subsystem of the processing resources of the allocentric condition.