The current study aimed to replicate a recent study conducted by Narayan, Mak, & Bialystok (2016) that found effects of top-down linguistic information on a talker discrimination task by comparing four conditions: compounds (day-dream), rhymes (day-bay), reverse compounds (dream-day), and unrelated words (day-bee). The original study used both within- and across-gender pairs and same and different trials were analyzed separately, obscuring possible response biases. Narayan et al. found graded performance between the four conditions, but some results were likely to have been influenced by the use of across-gender trials in the different-talker condition. In the current study, only female speakers were used and results were analyzed with signal detection theory (sensitivity and bias measures). Results revealed that participants were faster to respond to rhyming pairs than the three other conditions. In addition, participants were significantly more sensitive to talker differences in rhyming pairs than unrelated pairs, but no other conditions differed. Participants were more biased to respond “same” in the rhyme and compound conditions than in the unrelated condition. These results demonstrate a partial replication of the Narayan, Mak, & Bialystok (2016) findings, suggesting an interaction between linguistic and talker information during speech perception.