Due to a variety of reasons, one of the most common types of training found at companies is safety and health training. As part of a comprehensive health and safety training program there is usually an ergonomics training course. These courses are used to empower the employees to identify hazards and set up their workstations with the goal of injury prevention and increasing employee efficiency. Even with their usage, little data exist on the effectiveness of ergonomics training. In addition, no published research is available on the effectiveness of office ergonomics delivered via the web. This research project investigated the effectiveness of office ergonomics training delivered by both an instructor and a web-based program. Using a methodology popularized by Kirkpatrick, this investigation focused on the effects of both training delivery methods for knowledge, behavior, and reaction to training. As a method for comparing results, data was collected for both the knowledge and the behavior prior to and post-training delivery. Data for reaction to training was collected post training. This investigation used multiple methods of comparisons between base pre and post-training data and between the two training delivery methods. These methods include intra-group, inter-group, gain-score, and normalized-scores comparisons. The result form these comparisons showed that for both delivery methods there was a significant increase for knowledge and behavioral changes. Additionally, the group that received web-based training had a significantly greater increase for both behaviors and knowledge. However, there was no difference between the two training methods for reaction to training. For the study population assessed, this investigation shows evidence that both instructor and web-based office ergonomics training is effective at generating behavior change and knowledge gain. However, this study shows that web-based training was more effective at generating a greater change than the instructor delivered course. Additionally, this study provides evidence that the common method of assessing participate reaction to training is not effective at determining the true effectiveness of the training.