Much of our knowledge about the physiologic role of androgens in women is based on measurements, primarily in serum, using radioimmunoassay (RIA) methodology that involves purification of the analyte by organic solvent extraction and column chromatography. Although the extraction/chromatographic RIA is highly reliable when properly validated, it is time consuming and costly. For this reason, direct RIA methods were developed. Subsequently, the radioactive marker was replaced by other labels that have been used in direct chemiluminescent, fluorescent, and enzyme immunoassays on autoanalyzers. Although direct immunoassays are simple and rapid, they are seldom thoroughly validated, and generally lack the sensitivity and specificity for reliable measurements of testosterone levels found in postmenopausal serum. In recent years there has been increased use of mass spectrometry assay methods to quantify steroid hormones. These methods are touted to become the gold standard for all steroid hormone measurements. Because urine contains predominantly glucuronidated androgens, multistep procedures are required for the measurement, which is not practical for diagnostic testing. Androgens also can be measured in saliva, but major methodologic problems are associated with the measurements. In addition, there is a misconception that salivary testosterone levels reflect free testosterone levels in serum.