Serologists are making increasing use of the term "reproducibility" when referring to the reliability or repeatability of serological tests. The practical use of the concept, however, has been limited by the absence of an appropriate numerical scale on which differing reproducibilities can be quantified and objectively compared. This limitation can be overcome by adopting the proposed quantitative measure of reproducibility. The recommended measure is a natural extension of the common practice of considering a serological test to be acceptably reporducible so long as replicate titers remain within a twofold range. The measure can be readily used in the field of serology, and examples are given of how this can be done.