Examines the utility of choice response time (CRT) as a dependent variable in speech discrimination research. 12 undergraduate females responded to lists in which initial consonant phonemes differed by 1, 2, 3, or 4 Miller-Nicely distinctive features (DFs). 1/2 of the pairs in each list had the same initial consonants. No difference was found across lists. Mean CRT for identical pairs was significantly higher than for contrasting pairs. Total error rate was 1%. Posterior analysis revealed significantly more errors for the 1-DF difference list. Results suggest that CRT is an informative dependent variable for speech discrimination research. (PsycINFO Database Record (c) 2006 APA, all rights reserved). © 1969 American Psychological Association.