Exercising in cold environments results in water losses, yet examination of resultant voluntary water intake has focused on warm conditions. The purpose of the study was to assess voluntary water intake during and following exercise in a cold compared with a warm environment. Ten healthy males (22 ± 2 years, 67.8 ± 7.0 kg, 1.77 ± 0.06 m, VO₂peak 60.5 ± 8.9 ml·kg⁻¹·min⁻¹) completed two trials (7-8 days). In each trial subjects sat for 30 min before cycling at 70% VO₂peak (162 ± 27W) for 60 min in 25.0 ± 0.1 °C, 50.8 ± 1.5% relative humidity (RH; warm) or 0.4 ± 1.0 °C, 68.8 ± 7.5% RH (cold). Subjects then sat for 120 min at 22.2 ± 1.2 °C, 50.5 ± 8.0% RH. Ad libitum drinking was allowed during the exercise and recovery periods. Urine volume, body mass, serum osmolality, and sensations of thirst were measured at baseline, postexercise and after 60 and 120 min of the recovery period. Sweat loss was greater in the warm trial (0.96 ± 0.18 l v 0.48 ± 0.15 l; p < .0001) but body mass losses over the trials were similar (1.15 ± 0.34% (cold) v 1.03 ± 0.26% (warm)). More water was consumed throughout the duration of the warm trial (0.81 ± 0.42 l v 0.50 ± 0.49 l; p = .001). Cumulative urine output was greater in the cold trial (0.81 ± 0.46 v 0.54 ± 0.31 l; p = .036). Postexercise serum osmolality was higher compared with baseline in the cold (292 ± 2 v 287 ± 3 mOsm.kg⁻¹, p < .0001) and warm trials (288 ± 5 v 285 ± 4 mOsm·kg⁻¹; p = .048). Thirst sensations were similar between trials (p > .05). Ad libitum water intake adjusted so that similar body mass losses occurred in both trials. In the cold there appeared to a blunted thirst response.