The activity of trail running is described as a form of running sport taking place on the outdoor and mountainous environment, often incorporating hilly inclines and declines, with races lasting over 24 h. Event logistics, with emphasis on resource management, course logistics, and safety measures, are challenges to the organizations, where the planning of the arrival of runners could improve the management efficiency. The use of three velocity algorithms and Random Forest and Lasso machine learning algorithms were applied and measured to predict the arrival time of athlete at each checkpoint, analyzing the performance of the algorithm under runner or checkpoint perspectives, based on a well known event data. The case study results present an error reduction of 24.45% of error with the use of Random Forest machine learning algorithm over other algorithms, with a direct comparison against linear regression methods.