Students' frequent exposure to inquiry-based science teaching is presumed more effective than their exposure to traditional didactic instruction in helping improve competence in content knowledge and problem solving. Framed through theoretical perspectives of inquiry-based instruction and culturally relevant pedagogy, this study examines this assumption using TIMSS 2007 US eighth-grade student data. Data analysis revealed three instructional approaches, including more inquiry based, mixed, and practice based, were popularly practiced in classrooms. More inquiry-based instruction was not significantly associated with content and problem solving achievements across Caucasian, African American, and Hispanic American students. The mixed teaching approach had a significant positive relationship for Hispanic American student's content and problem solving achievement. The relationship between practice-based approach with content and problem solving achievements of all three racial/ethnic groups were not significant. This study questions the popular assumption of the effectiveness of science teaching approaches and offers important implications for science teaching reform.