The diagnosis of iron deficiency using the current commonly used tests is usually difficult in hemodialysis patients. Soluble transferrin receptor (sTfR) has caught the attention of physicians recently as regards its use as a parameter for the evaluation of iron status. This study was conducted in order to evaluate the correlation of serum soluble transferrin receptor (sTfR) concentration with hematological parameters and iron profiles, in the role of identifying iron deficiency among dialysis patients. Seventy-three patients having received chronic hemodialysis and stable maintenance recombinant human erythropoietin (rHuEPO) therapy were included. Iron, total iron-binding capacity, ferritin and sTfR were measured in the first week. Following this, these patients began to receive intravenous iron dextran (2 mg/kg/week) for 4 weeks. The hematocrit (Hct), hemoglobin (Hb) levels and reticulocyte counts were evaluated weekly. At the beginning of fifth week, the sTfR level was measured again. Patients were classified as belonging to one of the following groups: serum ferritin < 100 microg/L - absolute iron-deficient group; initial ferritin level > or = 100 microg/L with an increase in hemoglobin of greater than 1 g/dL at the end of the study occult iron deficiency group; others - non iron-deficient group. Seventy-one patients completed the study. The concentration of sTfR was positively correlated with Hct, Hb and reticulocyte index at the beginning (r = 0.236, p = 0.047; r = 0.257, p = 0.04; r = 0.401, p < 0.01, respectively) and at the end of the study (r = 0.384, p < 0.01; r = 0.338, p < 0.01; r = 0.427, p < 0.001, respectively). After 4 weeks of iron and rHuEPO therapy, the sTfR concentration increased, rather than declined, from 21.85 +/- 8.06 nM to 23.76 +/- 7.42 nM (p = 0.04) and the change was positively correlated with the changes in Hct, Hb and reticulocyte index. The administered rHuEPO doses did not differbetween the iron deficiency group (absolute deficiency, n = 3; occult deficiency, n = 10) and non-iron deficiency group (n = 58). The sTfR levels failed to identify the occult iron deficiency group because there was no difference between occult iron-deficient and non-iron-deficient patients (24.73 +/- 9.09 nM versus 21.60 +/- 7.89 nM, p = 0.34). Instead, transferrin saturation (TS) could be a differential marker between the 2 groups (19.0 +/- 10.9% versus 30.1 +/- 12.7%, p = 0.012). The serum sTfR concentration is indeed an appropriate marker for erythropoiesis. The erythropoitic effect of administered rHuEPO could mask the effect of iron status on the sTfR concentration. This might make the sTfR concentration no longer an appropriate index to identify the presence of occult iron deficiency. Thus, TS and ferritin currently remain better methods for the evaluation of iron status in rHuEPO-treated chronic hemodialysis patients.