OBJECTIVE Biochemical measurement of the hepatic iron concentration (HIC) is essential for the diagnosis of hereditary hemochromatosis (HH). The aim of this study was to determine whether the HIC at the time of diagnosis could predict the subsequent phlebotomy requirements and to determine whether saturation of HIC occurred in HH. METHODS Fifty-four patients (32 male, 22 female) with homozygous HH were evaluated, and HIC was measured in liver biopsies. Patients were subjected to weekly phlebotomy (500 ml) until the transferrin saturation was <50% and/or the serum ferritin concentration was <50 microg/L. The relationship between HIC and total body iron stores (as measured by phlebotomy requirements) was determined using both linear and nonlinear (sigmoidal model) least squares regression. RESULTS The HIC ranged from 3,742 to 41,040 microg/g dry wt. A linear relationship between HIC and total body iron stores (iron removed, IR, g) best described the data both in male (HIC = 1986 IR - 3494; r = 0.83; p < 0.001) and female HH patients (HIC = 1251 IR + 2690; r = 0.75; p < 0.001). Men required eight more phlebotomies (2 g iron) on average, compared with women, to reach normal iron stores. There was no evidence of saturation of hepatic iron levels at higher total body iron stores. However, accurate prediction of individual phlebotomy requirements based on the HIC or serum ferritin concentration at the time of diagnosis was not possible. CONCLUSIONS The phlebotomy requirement for treatment of HH cannot be accurately predicted from the initial HIC or serum ferritin level. Within the range examined, hepatic iron deposition did not saturate in HH.