Abstract
All immunoassays for female serum testosterone give falsely high results in some samples. The effect is variable and cannot be predicted for any given sample. Inaccurate calibration or interference by cross-reacting substances is almost certainly the cause of the problem, but for many immunoassays, the exact nature of the interferent is not known. Some of the interference can be removed by employing an extraction step prior to immunoassay. The advent of fast simple and sensitive liquid chromatography tandem mass spectrometry methods offers an exciting alternative to immunoassay for serum testosterone measurement. It is recommended that all high serum testosterone concentrations in women are checked, before reporting, by a method which is accurate (i.e. minimal bias to isotope dilution gas chromatography mass spectrometry [ID-GCMS] method) and is not subject to interference. Action should also be taken by assay users, manufacturers, regulators and professional bodies to ensure accurate standardization and comparability of assays.