Soluble transferrin receptor (sTfR) and ferritin concentrations were measured in a variety of clinical settings to compare the ability of these two tests to identify iron deficiency. Among 62 anemic patients who either had a bone marrow aspirate performed or had a documented response to iron therapy, the diagnostic sensitivity and specificity of sTfR (at a diagnostic cutoff of > 2.8 mg/L) were 92% and 84%, respectively, with a positive predictive value of 42% in this population. Ferritin (< or = 12 microg/L) had a sensitivity of 25% and a specificity of 98%. However, the sensitivity and specificity of ferritin could be improved to 92% and 98%, respectively, by using a diagnostic cutoff value of < or = 30 microg/L, resulting in a positive predictive value of 92%. Ferritin and sTfR were also measured in 267 outpatient samples and 112 medical students. In the outpatient group, the two tests agreed in 73% of the samples; however, 25% of the samples had ferritin values > 12 microg/L and increased sTfR. Among the medical students, there was 91% agreement between the two tests, but 7% of the samples had ferritin < or = 12 microg/L and normal sTfR. Together, these data suggest that measurement of sTfR does not provide sufficient additional information to ferritin to warrant routine use. However, sTfR may be useful as an adjunct in the evaluation of anemic patients, whose ferritin values may be increased as the result of an acute-phase reaction.