The current study investigates the influence of L2 proficiency on cognitive control among three matched groups of unbalanced Chinese-English bilinguals. Flanker task was administered to measure conflict monitoring and inhibition, and Wisconsin Card Sorting Test (WCST) to measure mental set shifting. ANOVA analyses of the Flanker results showed no differences in inhibition across all groups and no interaction between group and condition. However, the Flanker results showed faster performance for the highest L2 proficiency group relative to the lowest L2 proficiency group in all conditions (incongruent, neutral, and congruent), which reflects better ability of conflict monitoring. Finally, ANOVA analyses of the WCST results showed no differences across all groups. These results altogether suggest that L2 proficiency has significant influence on cognitive control, but only in conflict monitoring, not in inhibition or mental set shifting.