In the present study, the main objective was to study factors affecting postexercise voluntary water intake in horses. Four Standardbred horses (mean ± s.e. bwt 500 ± 8 kg) were used to study water intake and effects of altering hydration status before an incremental exercise test (INCR) and a 40 min constant velocity exercise test (CONST) on a treadmill. Exercise was performed during normohydration (N), after dehydration for 24 h (DEH) and after hyperhydration with 12 l water 30 min before exercise (HH). DEH resulted in a bodyweight loss of 3% and there were signs of some fluid uptake prior to exercise in both HH trials. By the end of the INCR, the calculated change in plasma volume (PVcalc) was −13 ± 1, −21 ± 1 and −11 ± 3% in the N, DEH and HH trials, respectively. During the highest exercise velocities a hypotonic shift of fluid was seen in all INCR trials. There was a greater accumulation of plasma lactate (pLA) in HH-than in N-INCR, probably caused by the extra weight to be carried. CONST induced a similar fluid loss (3%) in all trials, but the decrease in PVcalc at the end of exercise was significantly smaller in HH (−7 ± 2%) than in N (−14 ± 1%) and DEH (−19 ± 2%). In DEH-INCR and DEH-CONST, plasma sodium concentration (pNa) was higher than in N until drinking water was offered 1 h postexercise. In the presence of both an increased pNa and a decrease in PVcalc when dehydrated, the horses drank immediately when offered water postexercise. In N-CONST, there was a significant decrease in calculated PVcalc (−10 ± 2%) but no increase in pNa when water was given and in this trial the horses rehydrated less rapidly. Plasma aldosterone concentration (PAC) had increased to the same magnitude in all trials afterabout 10 min, irrespective of type of exercise or hydration status. It was concluded that when both an osmotic and hypovolemic thirst stimulus was present, the horses rehydrated more rapidly postexercise.