The quantification of compounds using isotope dilution mass spectrometry re
quires the establishment of calibration curves prior to determination of an
y unknown sample. When calibration over a wide concentration range is requi
red and/or when an overlap exists between internal standard and analyte ion
s (if mono- or di-isotopically-labelled internal standards are used), secon
d-order calibration curves are obtained. In this paper we have compared sev
eral calculation methods to linearize such calibration curves. We found tha
t the method published by Bush and Trager(6) gives a satisfactory linear re
lationship between the corrected amount ratio y = Q(1)(Q(u) + tQ(1)) (the v
alue Q(u) being the amount of unlabelled analyte, Q(1) the amount of labell
ed internal standard and t, the fixed fraction of the internal standard, wh
ich is identical to the unlabelled analyte) and the ratio of unlabelled to
labelled ion intensities. All the other calculation methods that have been
published so far have failed to linearize the second-order calibration curv
e build-up over a wide concentration range. Copyright (C) 2001 John Wiley &
Sons, Ltd.