My project focused on using a supervised machine learning approach to predict domestic flight delays 2 hours and 24 hours prior to departure. Models were trained on flight and weather data for 348 major airports in the United States between January 2020 and December 2022. An iterative approach to feature engineering, feature selection, hyperparameter tuning, and model comparison was used to maximize prediction accuracy. After this iterative process feature importance was calculated to determine the factors with the greatest impact on flight outcome. Conclusions were then outlined on what variables could be controlled to minimize risk of flight delays as well as recommendations for future research.