An immunoassay's minimal detectable concentration ((MDC), the smallest
analyte concentration the assay can reliably measure, is one of its m
ost important properties. Bayes' theorem is used to unify the five cur
rent mathematical MDC definitions. The unified definition has signific
ant implications for defining positive results for screening and diagn
ostic tests, setting criteria for immunoassay quality control and opti
mal design, reliably measuring biological substances at low concentrat
ions, and, in general, measuring small analyte concentrations with cal
ibrated analytic methods, As an illustration, we apply the unified def
inition to the microparticle capture enzyme immunoassay for prostate-s
pecific antigen (PSA) developed for the Abbott IMxm(TM) automated immu
noassay system. The MDC of this assay as estimated by our unifying app
roach is shown to be 4.1-7.1 times greater than currently reported, As
a consequence, the ability of the assay to measure reliably small con
centrations of PSA to detect early recurrences of prostate cancer is p
robably overstated.