Patients in maintenance treatment for HFE hemochromatosis since at least one year will be included in a two year study period and randomized in two groups experimental and control group. Because proton pump inhibitors are widely used as chronic medication, and because they can significantly modify iron absorption, patients will be stratified according to the use of proton pump inhibitors and gender. A first bloodletting will be performed at inclusion with the same volume as usually performed by the patients. Results of the biological test performed at this visit will guide the time schedule and volume of the next bloodletting according to randomization group (patients treated with bloodletting according to current guidelines ferritin alone versus patients treated with bloodletting according to transferrin saturation and serum ferritin).Blood count and iron metabolism parameters will be performed at each bloodletting and follow-up visits. Time schedule and volume of bloodletting will be adjusted to biological results after each follow-up visit. Volume and schedule for bloodlettings will be determined according to guidelines specifically designed for this study to assure harmonization of treatment management, and centrally validated through the recording of the biological tests in the electronic Case Report Form which will provide the investigator with the volume and schedule of the next bloodletting. There will be two ways of treatment modification: either change of schedule or volume of bloodletting. Patients will undergo follow-up visit every six months with clinical examination, questionnaires at J0, M12 and M24 (SF-36; AIMS2-SF, WOMAC, EQ-5D-5L), and biological test. For health economics analysis, data will be obtained thanks to a dedicated extraction from SNDS database SNDS database will allow to gather hospital stays, visits, and other healthcare-related costs as well as vital status (date (month/year) of death) and cause of death. A de-identified copy of the clinical database, restricted to the relevant variables, will be sent for semideterministic matching purpose with SNDS extraction using four key variables: gender, same date (month/year) of birth, same date (day/month/year) of visit for bloodletting; pending, of course, regulatory authorization.