Background: Although altered mental status is a term commonly used for older patients, the limits of the definition are unclear. Objectives: We aimed to determine the predictive factors of altered mental status and to standardize this broad-spectrum definition. Methods: The level and content of consciousness were assessed both in the basal status and in the emergency department presentation status of patients who were aged 65 years and above. The Richmond Agitation-Sedation Scale (RASS) for level and five features for content of consciousness were used to determine the actual change in consciousness. Results: Among 1250 patients, the rate of true altered mental status was 7.7% (n = 96), transient altered mental status was 3.5% (n = 44), and non-altered mental status was 88.8% (n = 1110). The 1- and 3-month mortality rates were higher for patients with true altered mental status (32.3% and 40.6%, respectively) than for other patients (3.3% and 8.4%, respectively; p < .001). The following criteria predicted true altered mental status with a high specificity: a change in RASS score > 2 points, newly disorganized thoughts, a perception disorder, inattention and disorientation, and an inability to communicate. Conclusion: True altered mental status criteria can be used to determine alterations in consciousness that lead to high mortality, and they can also be used as a common language for current and further investigations.