Although researchers first identified the fat-soluble vitamin cholecalciferol almost a century ago and studies have now largely elucidated the transcriptional mechanism of action of its hormonal form, 1alpha,25-dihydroxyvitamin D(3) [1alpha,25(OH)(2)D(3)], we know surprisingly little about mechanisms of vitamin D toxicity. The lipophilic nature of vitamin D explains its adipose tissue distribution and its slow turnover in the body (half-life approximately 2 mo). Its main transported metabolite, 25-hydroxyvitamin D(3) [25(OH)D(3)], shows a half-life of approximately 15 d and circulates at a concentration of 25-200 nmol/L, whereas the hormone 1alpha,25(OH)(2)D(3) has a half-life of approximately 15 h. Animal experiments involving vitamin D(3) intoxication have established that 25(OH)D(3) can reach concentrations up to 2.5 mumol/L, at which it is accompanied by hypercalcemia and other pathological sequelae resulting from a high Ca/PO(4) product. The rise in 25(OH)D(3) is accompanied by elevations of its precursor, vitamin D(3), as well as by rises in many of its dihydroxy- metabolites [24,25(OH)(2)D(3); 25,26(OH)(2)D(3); and 25(OH)D(3)-26,23-lactone] but not 1alpha,25(OH)(2)D(3). Early assumptions that 1alpha,25(OH)(2)D(3) might cause hypercalcemia in vitamin D toxicity have been replaced by the theories that 25(OH)D(3) at pharmacologic concentrations can overcome vitamin D receptor affinity disadvantages to directly stimulate transcription or that total vitamin D metabolite concentrations displace 1alpha,25(OH)(2)D from vitamin D binding, increasing its free concentration and thus increasing gene transcription. Occasional anecdotal reports from humans intoxicated with vitamin D appear to support the latter mechanism. Although current data support the viewpoint that the biomarker plasma 25(OH)D concentration must rise above 750 nmol/L to produce vitamin D toxicity, the more prudent upper limit of 250 nmol/L might be retained to ensure a wide safety margin.