Resting-state functional magnetic resonance imaging (rsfMRI) provides measurement of functional connectivity between brain regions. It represents a powerful tool to characterize whole-brain networks. However, several intrinsic issues have limited the potential of rsfMRI from being fully realized. In this presentation I will discuss these issues, and argue that they can be partially solved with rodent models, in which other neuroscience techniques such as optogenetics/chemogenetics can be used along with rsfMRI.