Two series of experiments were conducted to examine the effect of ingesting beverages with differing carbohydrate (CHO) concentrations and osmolalities on metabolism and performance during prolonged exercise in different environmental conditions. In series 1, 12 subjects performed three cycling exercise trials to fatigue at 70% VO2peak in either 33 degrees C (N = 6) (HT1) or 5 degrees C (N = 6) (CT). Subjects ingested either a 14% CHO solution (osmolality = 390 mosmol.1(-1) (HCHO); a 7% CHO solution (330 mosmol.1(-1) (NCHO) or a placebo (90 mosmol.1(-1) (CON1). In series 2, six subjects performed the same three trials at 33 degrees C (HT2), while ingesting either NCHO, a 4.2% CHO solution (240 mosmol.1(-1) (LCHO) or a placebo) (240 mosmol.1(-1) (CON2). Plasma glucose was higher (P < 0.05) in HCHO than NCHO, which in turn was higher (P < 0.05) than CON1 in both CT and HT1. Plasma glucose was lower (P < 0.05) in CON2 compared with NCHO and LCHO in HT2. The fall in plasma volume was greater (P < 0.05) in HCHO than other trials in both CT and HT1 but was not different when comparing the three trials in HT2. Exercise time was not different when comparing the trials in either HT1 or HT2 but was longer (P < 0.05) in NCHO compared with HCHO, which, in turn, was longer (P < 0.05) than CON1 in CT. These data demonstrate that, during prolonged exercise in the heat, fatigue is related to factors other than CHO availability. In addition, during exercise in 5 degrees C a 7% CHO solution is more beneficial for exercise performance than a 14% CHO solution.