The Diagnostic Speaker Recognizability Test (DSRT) is based on the principle that recognition of voices by human listeners presupposes discrimination with respect to various perceived voice traits (PVT's). With knowledge of the nature of such traits, we can evaluate the impact of speech degradation on speaker recognizability in terms of its effects on the discriminability of the various PVT's. Research over a period of 40 years has led to the identification of 20 PVT's. With the DSRT, listening crews rate five exemplars on each of the 20 PVT's. The effects of a system or device are measured as the loss of speaker identity information (SII) contained in such ratings relative to that contained in ratings made on undegraded speech samples.