The aim of this study was to determine the effect of carbohydrate (CHO) versus placebo (PLA) beverage consumption on the immune and plasma cortisol responses to a soccer-specific exercise protocol in 8 university team soccer players. In a randomized, counterbalanced design, the players received carbohydrate or placebo beverages before, during and after two 90 min soccer-specific exercise bouts (3 days apart) designed to mimic the activities performed and the distance covered in a typical soccer match. Blood and saliva samples were collected before, during and after the exercise protocol. Plasma lactate concentration increased to approximately 4 mmol x l(-1) at 45 and 90 min of exercise in both treatments (P<0.01). Plasma glucose concentration was significantly lower after 90 min of exercise with ingestion of the placebo than the carbohydrate (PLA: 4.57+/-0.12 mmol x l(-1); CHO: 5.49+/-0.11 mmol x l(-1); P<0.01). The pattern of change in plasma cortisol, circulating lymphocyte count and saliva immunoglobulin A secretion did not differ between the carbohydrate and placebo trials. Blood neutrophil counts were 14% higher 1 h after the placebo trial than the carbohydrate trial (PLA: 4.8+/-0.5x10(9) cells x l(-1); CHO: 4.2+/-0.5x10(9) cells x l(-1); P = 0.06), but the treatment had no effect on the degranulation response of blood neutrophils stimulated by bacterial lipopolysaccharide. We conclude that, although previous studies have shown that carbohydrate feeding is effective in attenuating immune responses to prolonged continuous strenuous exercise, the same cannot be said for a soccer-specific intermittent exercise protocol. When overall exercise intensity is moderate, and changes in plasma glucose, cortisol and immune variables are relatively small, it would appear that carbohydrate ingestion has only a minimal influence on the immune response to exercise.