Genetic selection for low emotional reactivity has been linked to a lower glucocorticoid response to exposure to stressors.  This study tested whether this link is due to changes in the activity of the HPA axis, by comparing resting diurnal activity and the responsiveness of the feed forward mechanisms of the HPA axis in Merino sheep that had been selectively bred for their high or low behavioural reaction to the stressors of isolation and human presence.  To study the resting diurnal activity of the HPA axis, we measured plasma glucocorticoidconcentrations of low and high emotionally reactive animals (n = 16) 24 h.  To test responsiveness of the pituitary-adrenal axis, we measured the glucocorticoid response to treatment with arginine vasopressin (0.0001 µg/kg B.W.), corticotrophin releasing hormone (0.01 µg/kg B.W.), or a mixture of both. To measure the responsiveness and sensitivity of the adrenal gland, we administered adrenocorticotrophic hormone (ACTH) in 4 small doses (0.0125 i.u., 0.05 i.u., 0.2 i.u., and 0.8 i.u.) and measured the speed and magnitude of the glucocorticoid response. There was no effect of temperament upon resting diurnal glucocorticoid concentrations, or upon the speed or magnitude of glucocorticoid response in any of our experiments.  Therefore, changes in the glucocorticoid stress response associated with emotional reactivity are not due to differences in the feed forward mechanisms of the HPA axis, but could be due to differences in the initial brain perception of the stressor, or changes in the negative feedback mechanisms of the HPA axis.