Sensory computations evolve as a function of internal state: are we hungry or satiated, sleepy or alert? To investigate how the brain implements this flexible processing of sensory information, we chose to study the larval zebrafish because of its well-defined behavioral repertoire and its optical accessibility for brain imaging. We recorded whole-brain activity at single cell resolution in freely moving fish while simultaneously presenting whole-field, dark-flash visual stimuli. Behaviorally, dark flashes reliably elicited high amplitude turn responses in awake and alert fish. Neurally, they drove robust visual responses in single neurons throughout the zebrafish pretectum, optic tectum, dorsal thalamus, and pallium. During quiescence, however, behavioral responses to the dark flash were abolished and many visual neurons showed striking concomitant changes in their visual response magnitude. Interestingly, these changes were highly variable; some neurons increased their response and others decreased their response. At the population level, correlation structure and inferred functional connectivity between the brain regions containing visually responsive neurons were also dependent on animal state. Thus, information flow between visually responsive brain regions is likely to be modulated by the animal’s locomotion state and may underlie the observed sensorimotor gating of behavioral responses to dark flashes. More broadly, this work helps frame the interpretation of qualitatively similar changes in population activity of visual neurons in mammalian systems, where it is currently only possible to record from a small fraction of the complete visual circuit.