<p>The present study systematically compared the neural processing of tones, vowels, and consonants in Mandarin syllable perception for both first language (L1) and second language (L2) listeners. Using a passive oddball paradigm, we evaluated participants’ syllable perception at a pre-attentive stage with mismatch negativity (MMN). Results confirmed that both L1 and L2 groups exhibited typical MMNs to consonant-, vowel-, and tone-deviant syllables due to their acoustic and/or phonological differences. Specifically, the L1 group exhibited similarly larger MMNs to the vowel- and tone-deviant syllables than to the consonant-deviant syllables, with vowels eliciting the earliest latencies, followed by tones, and consonants demonstrating the latest latencies. The L2 group exhibited graded MMN amplitudes and latencies across three deviation types, with vowels eliciting the strongest and earliest response, consonants an intermediate response, and tones the weakest and latest response. Group comparisons highlighted overall weakened and delayed MMNs for the L2 listeners than L1 listeners, in support of the reduced automaticity inherent in L2 perception. Taken together, our findings reveal the extent and efficiency of Mandarin sub-syllabic processing for both L1 and L2 listeners, and are particularly informative of the neural basis underlying the L2 tone-to-segmental disadvantage.</p>