Second-language learners (L2ers) weight cues (e.g., pitch) as a function of how the cues are used in the native language. This study investigates the effect of native dialect on the use of Chinese (i.e., Mandarin Chinese) tonal information by native speakers of Seoul Korean (SK) and Kyungsang Korean (KK). While SK does not use pitch to realize lexical prosody, KK uses pitch to realize lexically contrastive words. Intermediate-to-advanced SK-speaking or KK-speaking L2ers of Chinese (at the same Chinese proficiency) completed a forced-choice tone identification task and a speeded AX tone discrimination task. In the identification experiment, participants heard natural tonal stimuli carried by multiple syllables; in the discrimination experiment, the tonal stimuli were resynthesized to model on natural citation forms of two native speakers (one male, one female), and were superimposed on the vowel /y/. Data collection, with 15 SK listeners and 15 KK listeners to be tested, is ongoing in Shanghai. KK listeners are predicted to have a higher accuracy than SK listeners in identifying tones, and to weigh pitch contour more than SK listeners in MDS analyses of RTs when discriminating tones. If our hypothesis is correct, native dialect is suggested to be considered in L2 speech perception models.