Saliva, popular for the measurement of cortisol concentrations, can be easily and painlessly obtained, so that study participants or medical patients may collect their own samples. This raises the question of whether cortisol concentrations are stable if samples are mailed unfrozen. Seventeen adult subjects (five males, 12 females, mean age = 27.82, SD = 7.55) participated in this study. One saliva sample from each subject was split. Half were frozen within 1 h. The other was exposed to conditions that would mimic a postal trip, including wide variations in temperature and movement over 5 days. A statistically significant positive correlation between cortisol concentration in the frozen and nonfrozen saliva samples was found (R2 = 0.92, p < .001). A paired t-test revealed no significant difference between samples (t(16) = 1.56, n.s.). This indicates that cortisol concentrations are stable during extended periods without freezing when exposed to widely varying temperatures and movement.