The earliest national accountability tests occur at the fourth-grade level and the outcomes are aggregated. This national study exposes the depth of science education inequities a full year earlier by disaggregating third graders' demographic data with the science performance. A full 3-way ANOVA model was applied to test the 3 main effects (gender, ethnicity, and poverty level), 2-way interaction effects, and 3-way interaction effects on science performance. The 3-way interaction revealed additive benefits and consequences for students at the margins. At/above-poverty white males had the highest science performance, and below-poverty African American females had the lowest. These analyses suggest that science equity does not exist as early as third grade. Interventions in science education should begin earlier and address the inequities in science teaching, learning, and assessment related to gender, ethnicity, and poverty.