Airway management problems are key drivers for anesthesia-related adverse events. Awake tracheal intubation using flexible bronchoscopes with preserved spontaneous breathing (ATI:FB) is a recommended technique to manage difficult tracheal intubation in anesthesia, intensive care and emergency medicine. However, a prospective developed classification for this type of airway management is lacking. Due to the absence of a specifically tailored, validated classification for awake intubation with flexible bronchoscopes, many airway operators and institutions use classification tools that were originally developed for direct laryngoscopy, such as the percentage of glottic opening (POGO) score or Cormack-Lehane classification, although their diagnostic performance for the classification of ATI:FB is unknown. This prospective model development and validation study aims to develop two multivariable prediction models: a diagnostic prediction model to classify difficult ATI:FB after ATI:FB has been performed and a second prognostic prediction model to predict the risk for difficult ATI:FB before ATI:FB is performed. An additional aim is to develop a machine learning algorithm to evaluate ATI:FB.