Reduced levels of serum albumin concentration, a routine blood test, within the "normal" range have been reported to be associated with mortality risk. The literature is reviewed, with a focus on cohort studies meeting specified criteria, and findings are summarized. In studies of many populations, comprising healthy subjects and patients with acute or chronic illness, serum albumin concentration is inversely related to mortality risk in a graded manner over its entire range; the estimated increase in the odds of death ranges from 24% to 56% for each 2.5 g/l decrement in serum albumin concentration. The association predicts overall and cause-specific mortality including cardiovascular mortality. It is likely that albumin concentration is a highly sensitive indicator of preclinical disease and disease severity. A direct protective effect of the albumin molecule is suggested by the persistence of the association after adjustment for other known risk factors and preexisting illness, and after exclusion of early mortality. Although biologically plausible, there is no direct evidence for this hypothesis. Serum albumin concentration is an independent predictor of mortality risk and could be useful in the quantification of risk in a broad range of clinical and research settings.