This study evaluates the ability of large language models (LLMs) to assess clinical acuity in emergency department (ED) visits. The researchers selected 10,000 pairs of adult ED visits from January 1, 2012, to January 17, 2023, at the University of California, San Francisco, where each pair had different Emergency Severity Index (ESI) acuity levels. The LLMs were queried to identify the patient with higher acuity based on their clinical history. The LLMs demonstrated an accuracy of 89% in correctly inferring the higher-acuity patient, which was comparable to the performance of human physicians in a 500-pair subsample. The study suggests that integrating LLMs into ED workflows could enhance triage processes while maintaining quality, warranting further investigation. The findings highlight the potential of LLMs to streamline clinical decision-making and improve patient care.This study evaluates the ability of large language models (LLMs) to assess clinical acuity in emergency department (ED) visits. The researchers selected 10,000 pairs of adult ED visits from January 1, 2012, to January 17, 2023, at the University of California, San Francisco, where each pair had different Emergency Severity Index (ESI) acuity levels. The LLMs were queried to identify the patient with higher acuity based on their clinical history. The LLMs demonstrated an accuracy of 89% in correctly inferring the higher-acuity patient, which was comparable to the performance of human physicians in a 500-pair subsample. The study suggests that integrating LLMs into ED workflows could enhance triage processes while maintaining quality, warranting further investigation. The findings highlight the potential of LLMs to streamline clinical decision-making and improve patient care.