We are all familiar with the idea of automatic skills. Anyone who has learned to read, or drive a car, has developed some skills to such an extent that they can be executed with little thought or ongoing monitoring. Furthermore, we seem to be able to execute these skills and still have plenty of mental capacity left over to undertake other mental tasks. So, it is a seemingly trivial task to drive a car and think about taking a route that will avoid the traffic jam ahead. Trivial, that is, to an experienced driver. A novice would struggle with both tasks attempted separately, let alone simultaneously. Much research has been conducted on the concept of automaticity in the field of cognitive psychology. There are several theories that purport to explain what takes place in the mind as someone practices a skill for years until they reach the point of automaticity. Curiously, these same theories seem to ignore another important feature of skill acquisition; the transfer of skills. Imagine you had spent some time learning the game of tennis. How good would you then be at the game of squash? Admittedly the courts are quite different, the rules are very different, and so are the balls. Nonetheless, both games use similar racquets, and both require accurate eye-hand co-ordination, as well as speed and fitness. Several theories describe the conditions under which transfer between skills can occur, but none of these theories explain well what happens when a skill has become automatic. There also appears to be a paucity of research on the specific question of whether automaticity facilitates or hinders transfer. The Cognition Research Group has commenced several projects to examine this issue. This research has clear implications for education and training. For example, should we train people to be automatic on certain skills before we introduce higher order skills? Certainly some research suggests that overloading trainees too early in training can be counter-productive. On the other hand, can automaticity create skills that are too stereotypical and, hence limit transfer? We hope to provide answers to these questions and so produce guidelines for more effective training programs. These answers will also feed into another CRG project (Arithmetic Computer Game).