The righting reflex (RR) is frequently used to assess level of arousal and applied to animal models of a range of neurological disorders. RR produces a binary result that, when positive, is used to infer restoration of consciousness, often without further behavioral corroboration. We find that RR is an unreliable metric for arousal/recovery of consciousness. Instead, cortical activity and motor behavior that accompany RR are a non-binary, superior criterion that accurately calibrates and establishes level of arousal in rodents.