To date, research involving functional neuroimaging of typical and atypical development has depended on several assumptions about the postnatal maturation of the brain. We consider evidence from multiple levels of analysis that brings into question these underlying assumptions and advance an alternative view. This alternative view, based on an “interactive specialization” approach to postnatal brain development, indicates that there is a need to: obtain data from early in development; focus more on differences in interregional interactions rather than searching for localized, discrete lesions; examine the temporal dynamics of neural processing; and move away from deficits to image tasks in which atypical participants perform as well as typically developing participants.