Functional neuroimaging of the human brain in the resting state has revealed distinct networks associated with distinct cognitive functions. However, having parcellated the brain into functionally distinct areas, we are still only beginning to understand the dynamics of how networks interact on a systems-scale. I will present recent work from the Raichle lab exploring systems-level dynamics from the perspective of highly organized, reproducible temporal sequences in spontaneous fMRI data. We have demonstrated that not only are there stereotyped temporal sequences embedded in resting state fMRI, but also that these temporal dynamics are profoundly altered by arousal state and relate to human behavioral phenotypes. Finally, I will present new work illustrating how analysis of spontaneous temporal sequences in mouse imaging and electrophysiology allows us to further understand the underlying neurophysiology of systems-level communication in spontaneous low-frequency activity.