This inquiry explores whether second language learners’ stress processing can be clarified by the extent to which suprasegmental cues contribute to lexical identity in the native language. It centers on Standard Mandarin, Taiwan Mandarin, and American English listeners’ stress processing in English nonwords. In Mandarin, fundamental frequency contributes to lexical identity by signaling lexical tones, but only in Standard Mandarin does duration identify stressed–unstressed and stressed–stressed words. Participants completed sequence-recall tasks including English disyllabic nonwords contrasting in stress. Experiment 1 employed natural stimuli; Experiment 2 employed resynthesized stimuli that isolated fundamental frequency and duration cues. Experiment 1 disclosed no difference amid the groups; in Experiment 2, Standard Mandarin listeners employed duration more than Taiwan Mandarin listeners did. These results are interpreted within a cue-weighting theory of speech perception.