Intrinsic functional connectivity magnetic resonance imaging (fcMRI) has emerged as a powerful tool for mapping large-scale networks in the human brain. Robust and reliable functionally coupled networks can be detected in individuals that echo many known features of anatomical organization. Features of brain organization have been discovered, including descriptions of distributed large-scale networks interwoven throughout association cortex, interactions (including anticorrelations) between brain networks and insights into the topography of subcortical structures. But interpreting fcMRI is complicated by several factors. Functional coupling changes dynamically, suggesting that it is constrained by, but not fully dictated by, anatomic connectivity. Critically to study of between-group differences, fcMRI is sensitive to head motion and to differences in the mental states of participants during the scans. We discuss the potential of fcMRI in the context of its limitations.