The most frequently used design to assess change as a result of training is the pre-test, post-test method. While a useful and easily set-up procedure, the pre-then-post approach can reduce rather than reflect actual changes. In this brief article, the author outlines a different procedure, called the post-then-pre method. This method is described as being significantly more suitable for determining the magnitude of change. Although the specific example used in this article concerns training in peer assistance, the model can be used in virtually any training program where it is expected that participants will change skill, knowledge, or attitude as a result of the training.