CONTEXT AND IMPORTANCE:With over 300 million surgeries performed under general anaesthesia annually, optimising perioperative brain health has become a critical public health priority. Electroencephalogram (EEG) monitoring, initially conceived to prevent awareness during anaesthesia, is now emerging as a tool for assessing cognitive vulnerability and predicting neurocognitive outcomes. OBJECTIVES:This narrative review synthesises the evolution of intraoperative EEG monitoring across three conceptual stages: from a depth-of-anaesthesia measure to a tool for optimising anaesthetic exposure, and to an emerging biomarker of brain frailty and long-term perioperative risk. KEY FINDINGS:We review the technical foundations of EEG signal analysis and the distinctive patterns produced by GABAergic anaesthetics, characterised by frontal alpha anteriorization and slow-wave oscillations. EEG-guided anaesthetic titration reduces drug consumption, accelerates recovery, and minimises environmental impact, while maintaining equivalent safety. Specific intraoperative EEG patterns, including burst suppression and diminished alpha power, are robustly associated with postoperative delirium and cognitive decline, especially in vulnerable populations. However, trials targeting these patterns have not consistently prevented delirium, revealing that baseline cognitive impairment mediates most of the risk. Age-dependent EEG changes under anaesthesia, including reduced oscillatory power and altered spectral characteristics, may serve as markers of latent brain vulnerability. CLINICAL IMPLICATIONS:Intraoperative EEG monitoring should be viewed not merely as a tool for titrating hypnotic agents, but as a window into perioperative brain health. Future integration of machine learning algorithms and longitudinal cognitive follow-up may enable EEG-based preoperative risk stratification and individualised perioperative management strategies, transforming perioperative care for cognitively vulnerable patients.