Critical illness is associated with dramatic changes in the glucocorticoid axis. These changes occur at all levels of the axis, are most obvious in the increased serum levels of glucocorticoid but also occur within the tissue. Frank impairment of these responses (e.g. in Addisons disease or with drugs that block adrenal hormone secretion) is associated with an adverse outcome but it has recently been suggested that many critically ill patients develop subtle abnormalities of the glucocorticoid axis that can worsen their illness. This has led to the development of new concepts such as functional adrenal insufficiency, relative adrenal insufficiency and corticosteroid insufficiency. How critical illness causes these abnormalities is unclear but proposed mechanisms include impairment of adrenal corticosteroid synthesis and cytokine related glucocorticoid resistance. Recent studies have attempted to define clinically useful measures of dysfunctional glucocorticoid responses that might be used to guide treatment. These have generally assumed that the abnormality in the axis is at the level of adrenal but this may not be the case. The current focus on circulating levels of glucocorticoid for instance cannot capture changes that happen with the tissue or cell. Which cells and tissues are most sensitive to changes in the glucocorticoid axis is currently unclear. Although supplemental glucocorticoids can improve biomarkers such as vascular tone it is uncertain whether this short-term benefit translates into long-term survival. Future research will need to clarify which tissues are important in the action of glucocorticoids and whether changes in glucocorticoid responsiveness vary between tissues. Additionally, improved ways to measure tissue glucocorticoid sufficiency are needed to guide clinical trials. A further problem is that corticosteroid insufficiency might vary with the duration and severity of critical illness and results obtained with one type of illness cannot be automatically assumed to apply in another.