The cheetah (Acinonyx jubatus) is highly endangered because of loss of habitat in the wild and failure to thrive in captivity. Cheetahs in zoos reproduce poorly and have high prevalences of unusual diseases that cause morbidity and mortality. These diseases are rarely observed in free-ranging cheetahs but have been documented in cheetahs that have been captured and held in captive settings either temporarily or permanently. Because captivity may be stressful for this species and stress is suspected as contributing to poor health and reproduction, this study aimed to measure chronic stress by comparing baseline concentrations of fecal corticoid metabolites and adrenal gland morphology between captive and free-ranging cheetahs. Additionally, concentrations of estradiol and testosterone metabolites were quantified to determine whether concentrations of gonadal steroids correlated with corticoid concentration and to assure that corticosteroids in the free-ranging samples were not altered by environmental conditions. Concetntrations of fecal corticoids, estradiol, and testosterone were quantified by radioimmunoassay in 20 free-ranging and 20 captive cheetahs from samples collected between 1994 and 1999. Concentrations of baseline fecal corticoids were significantly higher (p = 0.005) in captive cheetahs (196.08 +/- 36.20 ng/g dry feces) than free-ranging cheetahs (71.40 +/- 14.35 ng/g dry feces). Testosterone concentrations were lower in captive male cheetahs (9.09 +/- 2.84 ng/g dry feces) than in free-ranging cheetahs (34.52 +/- 12.11 ng/g dry feces), which suggests suppression by elevated corticoids in the captive males. Evidence for similar sulppression of estradiol concentrations in females was not present. Adrenal corticomedullary ratios were determined on midsagittal sections of adrenal glands from 13 free-ranging and 13 captive cheetahs obtained between 1991 and 2002. The degree of vacuolation of cortical cells in the zona fasciculata was graded for each animal. Corticomedullary ratios were larger (p = 0.05) in captive cheetahs; however, there was no difference (p = 0.31) in the degree of corticocyte vacnolation between the two populations. These data proxile both mnorphologic and functional evidence suggestive of chronic stress in captive cheetahs. Further research into the role of hypercortisolemia in the pathogenesis of the reproductive abnormalities and unusual diseases of captive cheetahs is needed.