The small-dose (20 mg) oral iron absorption test (OIAT) was performed in 76 hospitalized elderly patients and 30 healthy adults. Of the elderly patients, 34 were considered as iron deficient (serum ferritin level < 20 micrograms/L) of whom 23 were anaemic and 11 not anaemic, 21 had the anaemia of chronic disorders (ACD) and another 21 were non-anaemic patients with a normal serum ferritin level. There was a significant inverse correlation between the serum ferritin level as a measure of iron store and the maximum increase in serum iron during a 3-h test (Cmax), in the elderly as well as in the healthy adult group. A decision limit of 80 micrograms/dL for Cmax is a good discriminant between absent (serum ferritin < 20 micrograms/L) and adequate body iron stores. Sixty-eight per cent of the patients with a serum ferritin level < 20 micrograms/L but virtually none of the ACD patients, non-anaemic elderly inpatients with normal serum ferritin levels and healthy adults had a Cmax level > 80 micrograms/L. Although further investigation is needed before the OIAT can be recommended as a valuable test for evaluating iron absorption, predicting mild iron deficiency and differentiating between different categories of anaemia, it seems worthwhile that more effort should be done to validate this simple and safe test.