The present study examined how native Mandarin listeners and native Korean listeners of L2 Mandarin perceive the Mandarin level tones (i.e., Tone 1 and Tone 3) and how the Mandarin tone identification interacts with the Korean lenis-fortis stop perception in native Korean listeners. Two perception experiments, namely Mandarin Tone 1-Tone 3 Identification Experiment and Korean Lenis-Fortis Identification Experiment, were conducted. Results showed that native Mandarin listeners identified the monosyllabic level tones mostly as Tone 1, whereas the level tones on the first syllable in disyllabic expressions were distinguished as either Tone 1 or Tone 3 in a categorical manner. It was also found that the proficiency of L2 Mandarin has a strong influence on non-native tone perception, in that the responses of native Korean listeners of advanced L2 Mandarin largely conformed with those of native Mandarin listeners in both monosyllabic and disyllabic conditions. By contrast, less proficient listeners' tone judgments in monosyllabic conditions considerably correlate with the pitch height change, while a tendency to categorically distinguish Tone 1 and Tone 3 emerged in disyllabic conditions, though with less well defined perceptual boundary. The present study further revealed that L2 tones play a role in the perception of L1 stops. Despite the fact that Mandarin stops resemble Korean fortis in VOT, native Korean listeners of proficient L2 Mandarin appeared to distinguish the lenis and fortis stops in a categorical fashion in the initial position of disyllabic expressions. A great deal of similarity found between the lenis-fortis identification and the Tone 1-Tone 3 identification indicates that F0 cues are facilitated for the lenis-fortis identification of native Korean listeners. Findings of this study, therefore, demonstrate the perceptual reorganization that VOT cues are overridden by F0 was particularly notable in native Korean listeners of proficient L2 Mandarin, while the perceptual cue weighting shift did not seem to occur in less proficient L2 listeners.