The purpose of this study was to determine if straight sprint training transferred to agility performance tests that involved various change-of-direction complexities and if agility training transferred to straight sprinting speed. Thirty-six males were tested on a 30-m straight sprint and 6 agility tests with 2–5 changes of direction at various angles. The subjects participated in 2 training sessions per week for 6 weeks using 20–40-m straight sprints (speed) or 20–40-m change-of-direction sprints (3–5 changes of 100°) (agility). After the training period, the subjects were retested, and the speed training resulted in significant improvements (p < 0.05) in straight sprinting speed but limited gains in the agility tests. Generally the more complex the agility task, the less the transfer from the speed training to the agility task. Conversely, the agility training resulted in significant improvements in the change-of-direction tests (p < 0.05) but no significant improvement (p > 0.05) in straight sprint performance. We concluded that straight speed and agility training methods are specific and produce limited transfer to the other. These findings have implications for the design of speed and agility training and testing protocols.