The effects of three isoenergetic diets on metabolic and appetite responses to prolonged intermittent walking were investigated. Eight male subjects mean age (±SD) 26 ± 3 years and maximal oxygen uptake ( VO2 max) 60 ± 4 ml.kg^-1min^-1, undertook three 450-min treadmill walks at intensities varying between 25-30% and 50-55% of VO2 max under three dietary conditions. In a balanced design the subjects were given breakfast, snacks and lunch containing total carbohydrate (CHO), protein (P) and fat (F) in the following amounts respectively (g/70 kg body mass): mixed diet, 320 CHO, 32 P, 88 F; high-CHO diet, 470 CHO, 52 P, 25 F; high-fat diet, 63 CHO, 44 P, 201 F. Energy expenditure and substrate oxidation were measured by indirect calorimetry and balance was calculated over the 450-min exercise period, including 45-min for lunch. Blood samples were taken before exercise and every 45 min during the exercise protocol to determine the metabolic and hormonal responses. The high-fat diet resulted in a negative total CHO balance (-82 ± 1g) and higher fat balance (-37 ± 33g) over the trial (high-fat vs. mixed and high-CHO diets, P < 0.05). The high-CHO meal resulted in significantly lower concentrations in the metabolites non-esterified fatty acids, glycerol, and 3-hydroxybutyrate (P< 0.05 vs. high-fat meal). The high-fat meal was accompanied by higher concentrations of plasma glucagon (P< 0.05 vs. high-CHO meal) and lower concentrations of plasma insulin (P < 0.05 vs. mixed and high-CHO meals) Subjective ratings of fatigue and appetite, using a visual analogue scale, showed no differences between the three trials. These results suggest that dietary manipulation can significantly alter substrate supply without compromising performance. The main source of energy in all trials was fat oxidation, although diet influenced the degree of total CHO and fat oxidation.