With high operational tempo and increasingly complex operational environments, the U. S. Army is increasingly using game-based training as a lower cost and more time-effective training method for both individual and collective training of tactical skills. However, there has been little empirical evidence to demonstrate the effectiveness of game-based training and to help leaders make decisions about their use. In response to the need for more evidence regarding the effectiveness of game-based training (GBT), an evaluation of training games supported by Virtual Battle Space 2 (VBS2): U.S. Army was conducted in operational settings. This report describes the methods, measures, and results of an evaluation with 165 Soldiers participating in GBT. Pre- and post-measures were administered that focused on measuring training effectiveness through individual level (e.g., task performance) and unit level (e.g., unit effectiveness) outcomes. Results demonstrate that, in general, the training influenced both individual (e.g., task performance) and unit level (e.g., unit cohesion) outcomes. In addition, situational characteristics (the amount the unit prepared for the training and the level of leader involvement during the training) influenced both types of outcomes.