It is well established that the ingestion of carbohydrate-containing drinks can improve the performance of prolonged exercise. The present study examined the effects of ingestion of water and two dilute glucose-electrolyte drinks on exercise performance and on cardiovascular and metabolic responses to exercise. Twelve subjects exercised to exhaustion on a cycle ergometer at a workload corresponding to 70% of maximum oxygen uptake (VO2,max) on five occasions each separated by 1 week. The first trial served to accustom subjects to experimental conditions. On one trial, no drinks were given and on the others subjects drank 100 ml every 10 min. Drinks consisted of water, an isotonic glucose-electrolyte solution (I: 200 mmol/l glucose; 35 mmol/l NA2; 310 mosmol/kg) and a hypotonic glucose-electrolyte solution (H: 90 mmol/l glucose; 60 mmol/l Na+; 240 mosmol/kg). Treatment order was randomized. Blood and expired air samples were taken and heart rate and rectal temperature measured at intervals during exercise. Median exercise time was greatest for treatment H (110.3 min) followed by treatment I (107.3 min), water (93.1) and no drink (80.7). Endurance times differed significantly overall, and for pairwise comparisons (P < 0.01) between the no-drink trial and both treatments H and I: a difference between water and no drink was seen at the 5% level. At exhaustion, a significant treatment difference was found for the change in plasma volume, with the greatest decrease (6.7%) on the no-drink trial and the smallest decrease (0.5%) on treatment H. Significant treatment effects were also observed for heart rate, rectal temperature and serum osmolality. The results suggest that the ingestion of glucose-electrolyte drinks can improve exercise performance even when the amount of added glucose is small, and that performance may also be enhanced, albeit to a lesser degree, by ingestion of water.