Increasing evidence suggests that success in science, technology, engineering, and math (STEM) fields is not only dependent upon one's actual STEM-relevant abilities but also upon one's STEM-relevant attitudes-in particular, math and spatial attitudes. Here, we examine whether simply mentioning the math or spatial relevance of a task affects children's performance and the moderating role of children's math and spatial attitudes. Further, we examine gender differences in performance given pervasive gender gaps in STEM and early-emerging gender differences in math and spatial attitudes. Participants (221 first- to fourth-grade children from the United States; 113 girls, 108 boys; 52% White, 16% Black, 14% Asian, 9% Hispanic or Latinx, 18% multiple races/ethnicities) were introduced to a novel task framed as tapping into math or spatial abilities (or no framing [control condition]). Children then completed math and spatial anxiety and self-concept measures. Results indicate that children who heard the math task framing were less accurate relative to children in the control condition, and the effect was larger for those with higher math anxiety or lower math self-concept, but it was not different for boys and girls. Children who heard the spatial task framing, however, performed comparably to children in the control condition. Though both math and spatial attitudes revealed identical patterns of gender differences (with higher anxiety and lower self-concept in girls than boys), there were no gender differences in performance. This study highlights the salient role of math attitudes early in development and provides key insights for future work aimed at increasing STEM outcomes. (PsycInfo Database Record (c) 2024 APA, all rights reserved).