This study evaluated the accuracy of ChatGPT in answering otolaryngology-related questions. A total of 93 questions were asked, including 21 patient-level questions and 72 clinical practice guideline (CPG)-based questions across seven otolaryngology domains. Answers were classified as correct, partially correct, incorrect, or non-answer based on compliance with CPGs or accepted medical knowledge. Correct answers were given in 45.5% of cases, with 71.4% of patient-level questions being correct and 37.3% of CPG-based questions correct. Partially correct answers were given in 31.8% of cases, with 28.6% of patient-level questions and 32.8% of CPG-based questions. Incorrect answers were given in 21.6% of cases, with 28.4% of CPG-based questions incorrect. Non-answers were given in 1.1% of cases. CPG-based questions were less likely to be correct than patient-level questions (p = 0.003). There was no difference in the rate of correct answers between CPGs published before or after the period of data collection cited by ChatGPT. The accuracy of ChatGPT responses varied by domain, with laryngology having the highest percentage of correct answers (60%) and head and neck oncology having the lowest (33.3%). The study found that ChatGPT provided correct answers to 45.5% of questions, incorrect answers to 21.6% of questions, and partially correct answers to 31.8% of questions. The accuracy of ChatGPT responses was lower for CPG-based questions compared to patient-level questions. The study concluded that while ChatGPT can provide some medical information, it is not reliable for accurate medical information, especially in the field of otolaryngology. Physicians and patients should understand the limitations of this software and consider using well-established journals and national content experts for more accurate information.This study evaluated the accuracy of ChatGPT in answering otolaryngology-related questions. A total of 93 questions were asked, including 21 patient-level questions and 72 clinical practice guideline (CPG)-based questions across seven otolaryngology domains. Answers were classified as correct, partially correct, incorrect, or non-answer based on compliance with CPGs or accepted medical knowledge. Correct answers were given in 45.5% of cases, with 71.4% of patient-level questions being correct and 37.3% of CPG-based questions correct. Partially correct answers were given in 31.8% of cases, with 28.6% of patient-level questions and 32.8% of CPG-based questions. Incorrect answers were given in 21.6% of cases, with 28.4% of CPG-based questions incorrect. Non-answers were given in 1.1% of cases. CPG-based questions were less likely to be correct than patient-level questions (p = 0.003). There was no difference in the rate of correct answers between CPGs published before or after the period of data collection cited by ChatGPT. The accuracy of ChatGPT responses varied by domain, with laryngology having the highest percentage of correct answers (60%) and head and neck oncology having the lowest (33.3%). The study found that ChatGPT provided correct answers to 45.5% of questions, incorrect answers to 21.6% of questions, and partially correct answers to 31.8% of questions. The accuracy of ChatGPT responses was lower for CPG-based questions compared to patient-level questions. The study concluded that while ChatGPT can provide some medical information, it is not reliable for accurate medical information, especially in the field of otolaryngology. Physicians and patients should understand the limitations of this software and consider using well-established journals and national content experts for more accurate information.