The effects of administering (1) 6L isotonic oral rehydration solution (ORS), similar in composition to plasma (except for an elevated potassium concentration) and with an osmotic skeleton and (2) 6L water (no osmotic skeleton), were evaluated in five thoroughbred horses following exercise-induced dehydration. The horses were exercised on a treadmill for 10 min at walk (1.7 m.s-1; approximately 15% VO2max), 40 min at trot (3.7 m.s-1; approximately 25% VO2max) and 10 min at walk (1.7 m.s-1; approximately 15% VO2max). Exercise was undertaken on a 3 degrees incline at 30 degrees C/80% RH. Solutions of water or ORS at 20 degrees C were administered by nasogastric tube over 60s 5 min following exercise. Mean weight loss following exercise was 9.2 +/- 1.7 kg (2.0 +/- 0.4% body weight; mean +/- SEM) with water and 9.2 +/- 1.1 kg (2.0 +/- 0.2% body weight) with ORS and was not different between treatments (P > 0.05). Water treatment resulted in a fall in plasma [Na+] (approximately 3 mmol.L-1) and C1- (1-2 mmol.L-1) concentrations by 30 min after administration and the effect persisted until the end of the study (300 min post fluids). There was little change in plasma total protein (TP) from that at the end of exercise, suggesting a failure of water to restore or maintain PV. In contrast, ORS administration resulted in a small increase in plasma [Na+] (1-2 mmol.L-1) and [Cl-] (2-3 mmol.L-1) with a corresponding decrease in plasma TP. By 120 min post ORS, plasma TP and PV were no longer significantly different from rest or pre-exercise (P > 0.05), whilst with water, TP was elevated (approximately 3-4 g.L-1) and PV reduced (approximately 4-5 mL.kg-1). Total urine output was not significantly different between water (1096 +/- 135 mL) and ORS (750 +/- 215 mL, P > 0.05). Estimates of expected plasma volume and electrolyte concentration changes as a result of either treatment compared well with measured changes of TP and PV. On the basis of calculated or measured changes, it was estimated that only 1L of water contributed to rehydration following exercise compared to 4L isotonic, plasma-like ORS. The administration of 6L ORS restored the PV deficit induced by exercise with minimal or no disturbance of plasma electrolyte concentrations. In contrast, water alone resulted in minimal improvement in PV. When fluid intake after periods of fluid loss, such as induced by exercise or transport, is not accompanied by food intake, the present study has clearly demonstrated that water alone is ineffective in promoting rehydration compared with an isotonic, plasma-like ORS.