The cause and mechanism of development of Alzheimer's disease (AD) remain unexplained. Hyperactivity of the hypothalamic-pituitary-adrenal (HPA) axis, denoted by adrenal cortisol hypersecretion, is a recognised feature of the condition but generally disregarded as causative, due to lack of association between AD and other hypercortisolemic states. However, a meta-analysis of published studies suggests a need for reappraisal. A specific circadian rhythm of cortisol hypersecretion pertains at mild-to-moderate AD stages, entailing increased levels at the circadian peak from a low nadir. This is in contrast to the continuously elevated levels that are characteristic of other hypercortisolemic states, e.g. Cushing's disease or major depression. This previously overlooked detail provides a starting premise here: that equating the form of hypercortisolism in AD with that in other states is inappropriate, as phasic and chronic elevation elicit different neuroendocrine effects. Theoretical implications are discussed in this review. Given the capacity of glucocorticoids and corticotropin-releasing hormone to induce AD-associated pathologies, I suggest a role for circadian cortisol hypersecretion in the initiation of sporadic AD; and propose a temporal mechanism for AD development featuring neuroinflammation-mediated suppression of central glucocorticoid receptor (GR) signaling. This latter may represent a critical phase in AD development, where the density of functional GR is proposed to underlie the “cognitive reserve”. Supporting evidence for this mechanism is drawn from the brain regional locations of AD neuropathologies, and from risk factors for AD development (aging, ApoE-4 genotype, and hypertension). Thus, it is argued that basal hypercortisolemia merits further scrutiny regarding AD causation and development.