Testosterone treatment of hypogonadal men improves their hemoglobin, but the mechanism is not understood.To investigate possible mechanisms by which testosterone stimulates erythropoiesis in hypogonadal older men with unexplained or iron-deficiency anemia.The Anemia Trial of The Testosterone Trials, a placebo-controlled study in older, hypogonadal men.Twelve academic medical centers.95 hypogonadal men (testosterone <275 ng/mL) ≥65 years with anemia (hemoglobin <12.7 g/dL). They were classified as having unexplained (58) or iron deficiency anemia (37).Testosterone or placebo gel for one year.Markers of iron metabolism during the first three months of treatment.Testosterone replacement significantly (p<0.001) increased hemoglobin in the 58 men who had unexplained anemia (adjusted mean difference 0.58 g/dL; 95% CI 0.31-0.85). Testosterone replacement tended to increase hemoglobin in the 37 men who had iron deficiency (0.38 g/dL; -0.19,0.95), but the response was more variable and not statistically significant (p=0.19). In men with unexplained anemia, testosterone replacement suppressed hepcidin [-8.2 ng/mL (-13.7, -2.7) p=0.004] and ferritin [-19.6 µg/L (-32.8, -6.3) p =0.004], but in men with iron deficiency, testosterone replacement did not. The decrease in hepcidin was moderately correlated with the increase in hemoglobin in the men with unexplained anemia (correlation coefficient -0.35, p = 0.01) but not in those with iron deficiency anemia (correlation coefficient -0.07, p = 0.73 ).Testosterone replacement of older hypogonadal men with unexplained anemia stimulates erythropoiesis associated with increased iron mobilization. This effect appears to be attenuated by iron deficiency.