The issue of whether bilinguals have advantages over monolinguals in cognitive functions has received ongoing research attention. Most researchers have agreed that continuously shifting between two languages enhances bilinguals' executive function, but several recent studies failed to find any evidence of bilingual advantage. In addition, the mechanism of bilingual advantage in executive function is not fully understood. Here, we hypothesized that a bilingual advantage should appear on tasks requiring an enhanced level of executive function, and tested this hypothesis in a non-language-based mixed culture context and single culture context. Proficient bilinguals and non-proficient bilinguals completed an Eriksen Flanker Task in these two contexts. The results showed that proficient bilinguals’ performance on incongruent trials was better than that of non-proficient bilinguals in the mixed cultural context, but not in the single cultural context. These findings cast important light on understanding the nature of bilingual advantage.