What should be the reference method for measurement of the hepatic iron concentration? In this issue of the American Journal of Hematology, Urru and colleagues [1] present further evidence that, in patients with thalassemia major and transfusional iron overload, chemical measurement of the iron concentration in a desiccated liver-biopsy specimen obtained by percutaneous biopsy is highly reproducible, as judged by measurement in a successive specimen, provided that two conditions are met. The first condition is the absence of cirrhosis, defined in this study as Ishak scores [2] of 5 or 6, because, with cirrhosis, the iron distribution within the liver is no longer uniform. The second condition is that the liver-biopsy specimen is adequate, 1 mg dry weight; variability increases with smaller samples. No complications were observed with liver biopsy, extending the record of safety of liver biopsy in expert hands [3][4][5]. These results corroborate previous observations from these and other investigators [5][6][7] that support chemical measurement of the iron concentration in a liver-biopsy specimen as the reference standard for measurement of the hepatic iron concentration. Still, dissonance in the details must be acknowledged; no standard method for liver biopsy procedure, specimen processing and analysis has been established. The biopsy needle (cutting needle or Menghini needle with saline flushing), exposure to saline, formalin, and fixation solutions, use of iron-free containers, specimen processing (fresh or paraffin-embedded; methods of deparaffinization and of digestion), methods of analysis (colorimetric, atomic absorption spectroscopy, inductively coupled plasma mass spectroscopy), and other factors differ among published studies and potentially influence results [5,6]. Hepatic iron concentrations measured in deparaffinized samples were a mean of 23% greater than those in desiccated tissue from fresh liver specimens [6]. The minimum weight for an adequate liver sample for analysis remains uncertain; evidence has been provided both for thresholds of 0.4 mg dry weight [8,9] and for 1 mg dry weight [1,4,5]. Reports of variability in the hepatic iron concentration in different areas of the liver have generally been restricted to studies of cirrhotic livers [10][11][12]; data are lacking in patients with transfusional iron overload without cirrhosis. Even so, the close relationship observed between the hepatic iron concentration and total body iron stores [4] in transfused patients with thalassemia who were free of cirrhosis suggests that the extent of variability is limited.An alternative view of the reference method for measurement of the hepatic iron concentration is that magnetic resonance imaging (MRI), recommended in a variety of guidelines, has become the de facto technique. Nonetheless, since acceptance of the manuscript by Urru et al.[1] for publication, a painstaking systematic review and meta-analysis of this use of MRI has appeared, [13] concluding that "measurements of liver iron concentration by ...