This study examined the effect of a pre-exercise meal with different glycaemic index (GI) and glycaemic load (GL) on immune responses to an endurance performance run. Eight men completed a preloaded 1 h run at 70 % VO2max on a level treadmill followed by a 10 km performance run on three occasions. In each trial, one of the three prescribed isoenergetic meals, i.e. high GI and high GL (H-H), high GI and low GL (H-L), or low GI and low GL (L-L) was consumed by the subjects 2 h before exercise. Carbohydrate intake (% of energy intake), GI, and GL were 65 %, 79·5, and 82·4 for H-H; 36 %, 78·5, and 44·1 for H-L; 65 %, 40·2, and 42·1 for L-L, respectively. The running time for the three trials was approximately 112 min at 70 % VO2max for the first hour and 76 % VO2max for the last 52 min. Consumption of pre-exercise high-carbohydrate meals (H-H and L-L) resulted in less perturbation of the circulating numbers of leucocytes, neutrophils and T lymphocyte subsets, and in decreased elevation of the plasma IL-6 concentrations immediately after exercise and during the 2 h recovery period compared with the H-L trial. These responses were accompanied by an attenuated increase in plasma IL-10 concentrations at the the end of the 2 h recovery period. The amount of carbohydrate consumed in the pre-exercise meal may be the most important influencing factor rather than the type of carbohydrate in modifying the immunoendocrine response to prolonged exercise.