Résumé : Background Anaemia is often multifactorial in the elderly, with a frequent association between iron deficiency anaemia (IDA) and anaemia of chronic disease (ACD). The primary objective of our study was to investigate whether baseline hepcidin measurement could be useful for identifying iron deficiency (ID) in anaemic elderly patients. The secondary objective was to assess whether baseline hepcidin concentrations correlated with the relative increase of transferrin saturation (TS) after an oral iron absorption test (OIAT). Methods Blood samples were collected between 7:30 am and 10:00 am in 328 geriatric outpatients, 102 underwent the OIAT. Types of anaemia were classified according biochemical and clinical criteria. TS and hepcidin were measured at baseline and 4 h after the iron dose. The ability of baseline hepcidin measurement to highlight ID in elderly anaemic patients was assessed using a receiver operator curve (ROC) analysis. Correlations between baseline hepcidin levels and the increment of TS following the OIAT were investigated using the Spearman coefficient. Results Among 328 included patients, 78 (23.8%) suffered from anaemia; 13 (4.0%), 19 (5.8%), 27 (8.2%) and 19 (5.8%) patients fulfilled criteria for IDA, IDA/ACD, ACD and unexplained anaemia, respectively. By multivariable analysis, creatinine, C-reactive protein, ferritin, Delta TS and Delta hepcidin were independently associated with baseline hepcidin concentrations. The area under the ROC curve (95% confidence interval) was 0.900 (0.830-0.970) for baseline hepcidin measurement. Baseline hepcidin levels correlated negatively with the relative increase in TS with a Spearman coefficient of -0.742. Conclusions Baseline hepcidin levels could be a useful tool to identify ID in anaemic elderly patients and may predict acute iron response following OIAT.