The application of sensitive metabolic tests, such as the deoxyuridine suppression test and measurement of homocysteine and methylmalonic acid, to cobalamin status has identified the entity of mild, preclinical cobalamin deficiency. This state, common in the elderly, responds to cobalamin therapy. Preclinical deficiency may exist within the nervous system as well, although this requires further study. Nevertheless, it is well to remember that not all low cobalamin levels and not all abnormal metabolite results reflect cobalamin deficiency. Interpretation of metabolic results still requires caution, as do proposals to raise the cut-off point for low cobalamin levels to capture some normal levels that are associated with metabolic abnormality. The recognition of mild, preclinical deficiency has opened up many important issues. These include identifying its causes, what should be done about it, and what the clinical impact of the hyperhomocysteinemia itself is. Although malabsorptive disorders, especially food-cobalamin malabsorption, underlie about half of all cases of preclinical deficiency, no cause can be found in the remainder of these cases; poor dietary intake appears to be uncommon. In addition, unusual states of neurologically symptomatic cobalamin deficiency are being recognized, such as nitrous oxide exposure in patients with unrecognized deficiency and severe deficiency in children of mildly deficient mothers. All of these have broadened and complicated the picture of cobalamin deficiency while providing greater opportunities for prevention.