This study investigated potential item bias in a large-scale grade 3 reading achievement test, specifically against culturally diverse students with limited familiarity with mainstream Canadian culture. Students were classified based on their first language and length of residence in Canada, which was used as a proxy for cultural familiarity. A multi-group differential item functioning (DIF) analysis revealed that, of the five items hypothesized by content experts to require a high level of cultural familiarity, three exhibited varying degrees of DIF across student subgroups. While the performance gap between groups tended to narrow with increased exposure to mainstream culture, significant differences persisted on items requiring substantial cultural familiarity, even among students who had resided in Canada for five years or more. These findings highlight the need to refine test development practices to ensure fairness in testing and provide more valid score interpretations for diverse populations, especially in culturally heterogeneous educational settings.