Hematocrit has been implicated in the triggering of cardiovascular events and the development of cardiovascular disease. Studies have demonstrated the reliability of hematocrit at rest, however, data are lacking about hematocrit during repeated acute stress exposures. The current study assessed the reliability of hematocrit during rest and stress. Hematocrit was measured in two sessions during rest and in response to mental, cold and exercise stress in 84 healthy men and women. The stress tasks consistently elicited increases in hematocrit. Absolute levels of hematocrit were highly reliable between testing sessions. Changes in hematocrit with exercise stress were reliable whereas changes associated with mental and cold stress were not reliable between sessions. The findings indicate that hemoconcentration during brief mental and physical stress is more reliable for absolute levels than change scores. The reliability of stress-induced hemocontration may be improved by more provocative challenges and repeated sampling during stress. © 2007 Elsevier B.V. All rights reserved.