Early scientists recognised the importance of foods as a source of energy for muscular contraction, observing that periods of starvation resulted in feelings of fatigue, lethargy and a difficulty even in standing (Mosso, 1891). The introduction of a reliable exercise ergometer and improved methods for expired air analysis during the early years of the 20th century meant that the study of the metabolic response to exercise gathered pace and the rates of carbohydrate and fat use at rest and during bouts of prolonged exercise were measured and shown to be affected by the composition of the preceding diet (Krogh & Lindhard, 1920). Subjects were reported to feel less fatigue after eating a diet high in carbohydrate. Debate about the interconversion of fat and carbohydrate continued for some years, however. Measurement of blood glucose concentrations in finishers of the Boston marathon showed the hypoglycaemia was present in some and carbohydrate ingestion was suggested as a remedy (Levine et al, 1924). The first study to definitively show beneficial effects on performance – albeit on only three subjects – was that of Christensen and Hansen (1939). The introduction of tissue biopsy techniques to human physiology in the 1960s enabled muscle glycogen utilization and storage to be directly measured, yielding new mechanistic insights into the relationship between diet, carbohydrate availability and fatigue (Bergstrom and Hultman, 1966; Bergstrom et al, 1967). This work spawned a vast body of literature examining both acute and chronic manipulation of dietary intake and its effects on pre-exercise preparation, performance and recovery. The overwhelming conclusion was that ingestion of carbohydrate prior to and during exercise was beneficial to endurance exercise. . In contrast, recent suggestions that a low-carbohydrate diet can benefit performance lack a strong evidence base and also have limited theoretical support. It has been known for more than a century that the energy available per litre of oxygen uses if greater for carbohydrate than for fat (Zuntz, 1901) and that the oxygen cost of exercise is less when carbohydrate is the substrate than when fat is being oxidised (Frentzel and Reach, 1901). Where oxygen availability is critical, therefore, carbohydrate should be the preferred fuel: it is difficult to see an advantage of increasing fat oxidation. The recent growth in the application of tracer and molecular techniques has enabled researchers to track the fate of ingested nutrients and better understand the signalling events that take place in skeletal muscle in response to exercise. Current work is characterising how alterations in substrate availability modulate many adaptive processes (Hawley et al, 2011), and these approaches will likely open the door to the design of personalised diet and exercise interventions to optimise training outcomes. A relatively recent development in exercise science has been the recognition that the fatigue process may depend more on events occurring within the brain than in peripheral tissues. Newsholme’s central fatigue hypothesis was an elegant attempt to describe a possible underlying mechanism that embraced the known metabolic changes occurring in the periphery (Blomstrand et al, 1988). This, however, was not a new idea, and Bainbridge wrote in 1919 that “There appear to be two types of fatigue, one arising entirely within the central nervous system, the other in which fatigue of the muscles themselves is superadded to that of the nervous system.” This seems a remarkable insight that remains largely unrecognised even today. Even earlier, Lagrange (1889) wrote that “”Fatigue is . . . a kind of regulator, warning us that we are exceeding the limits of useful exercise, and that work will soon become dangerous. Numerous physiological phenomena show us that the sensation of fatigue has its seat rather in the nerve-centres than in the muscles.” As with so much else in science, the concept of a “central governor” that regulates or limits exercise performance is not new.