Anemia is a frequent complication of cancer and its treatment. A defect in erythropoietin production has been advocated as being the main cause of anemia in cancer patients. We studied serum erythropoietin levels in 74 patients with solid tumors and in a control group consisting of 20 otherwise healthy individuals without any malignancy, who have only iron deficiency anemia. Serum erythropoietin levels were measured by enzyme immunoassay in cancer patients without anemia (n=34), and in anemic cancer patients (n=40); either receiving chemotherapy (n=21) or not (n=19). Anemic cancer patients were found to have decreased response of erythropoietin for a given hemoglobin level (mean, 40.1+/-34.7 u/ml), compared with the patients having only iron deficiency anemia (mean, 69.7+/-68.6 u/ml) (P<0.05). In patients with iron deficiency anemia having no malignancy, erythropoietin response was remarkably high and inversely correlated with the level of hemoglobin (r=-0.69; P=0. 05). Although there was no correlation between hemoglobin and erythropoietin response in cancer anemia (r=-0.07), serum levels of erythropoietin were found to be higher in anemic cancer patients (mean, 40.1+/-34.7 u/ml), compared with cancer patients with normal hemoglobin values (mean, 19.96+/-18.4 u/ml). There was not any statistically significant difference between erythropoietin levels in anemic cancer patients with or without chemotherapy (mean, 43. 7+/-37.7 u/ml and 41.9+/-30.08 u/ml respectively; P>0.05). No difference in serum erythropoietin levels were noted in patients treated with cisplatin or non-cisplatin containing regimens (mean, 48.36+/-33.12 u/ml and 38.55+/-43.52 u/ml, respectively; P>0.05). In this study, we demonstrated that anemia in cancer patients was caused by blunted erythropoietin response, rather than its quantitative deficiency. Serial measurements, however, should be considered in patients receiving chemotherapy.