Coding environmental information in spatial terms is a general propensity of our cognitive systems (e.g., the future is ahead of us, the past behind us). One prominent spatial coding is that of number. While it is now well established that adults’ processing of numerical information is accomplished by spontaneously deploying visuo-spatial mental resources, the specific nature, origins, and neural basis of this phenomenon are poorly understood. Accordingly, the present proposal aims to describe the format, function and development of the link between numbers and space. Following our preliminary findings of an intimate link between numerical magnitude and spatial length in infancy and childhood, we propose to further investigate the neural basis and nature of the number-space linkage through two general lines of experiments. Behavioral experiments on newborns and infants will investigate the format of the number-space linkage, by exploring central questions regarding the automaticity, specificity and symmetrical nature of the link. Neuroimaging experiments on adults will investigate the neural basis of the number-space linkage, by investigating the involvement of common neural populations in the coding of both sources of information, thus seeking anatomical evidence for the spatial format of number representation. The experiments are based on the hypothesis that numerical and spatial representations have a privileged link via common brain and cognitive systems that emerges in infancy and remains operative throughout adulthood. The project has immediate implications at a theoretical level for researchers on cognitive psychology, in particular on numerical cognition, as well as at an applied, educational level. Early mathematics skills is one of the best predictors of later academic performance, and one of the foundations of mathematics education is the understanding of number. Early mathematics instruction might be enhanced by building on the spatial nature of number.