Arousal state has a major impact on perceptual ability, task performance, and diverse aspects of physiology and behavior. During wakefulness, spontaneous fluctuations in arousal state strongly modulate neural activity in numerous brain regions in mice, but the lack of large-scale, deep imaging methods has prevented testing whether these fluctuations affect the entire brain uniformly. Moreover, it remains unclear whether spontaneous and sensory-evoked fluctuations in arousal state engage the same brain circuits or are fundamentally distinct processes. To address this gap, we used functional ultrasound imaging in awake, head-fixed mice and correlated the recorded whole-brain activity with pupil size fluctuations, known to track arousal. We characterized a large-scale ‘arousal pattern’ of brain regions that exhibited a distinct temporal dynamic. Next, we compared this spontaneous arousal pattern to the brain-wide pattern elicited by arousing stimuli (mild air-puffs) and found a large overlap between the two, the main difference being that external stimuli also activate sensory areas. Finally, we assessed whether arousing stimuli elicit the arousal pattern across different arousal states. We induced different arousal states by manipulating the tonic levels of noradrenaline, a well-established regulator of arousal, with optogenetics and pharmacology. We found that the cortical component, but not the subcortical component, of the arousal pattern was sensitive to the noradrenergic tone. Our work refines the role of arousal as a global modulator of neural activity by identifying a specific brain network that responds to spontaneous and evoked fluctuations in arousal, and by characterizing its constitutive components across sustained arousal states.