Although boys and girls start out performing similarly on science, technology engineering, and mathematics (STEM)-related school subjects, the gap between them widens as they mature, leading to the underrepresentation of women in STEM fields at the college level and in the workforce. One explanation for this phenomenon is stereotype threat: that females are intimidated by the persistent negative stereotypes about women in STEM fields. Examining the inception and development of stereotype threat among young children is crucial for understanding and bridging the STEM gender gap. Two experiments investigated the effect of stereotype threat on performance on a spatial task among children ages 4 to 10. After a gender identity activation manipulation, children replicated a series of designs with LEGO blocks. In Experiment 1 (N = 22), 4-yearold girls whose gender identity was activated were significantly slower than girls whose gender identity was not activated. In Experiment 2 (N = 160), boys and girls in kinder-garten through third grade did not perform differently in stereotype-threat conditions or a condition where gender was not activated. However, girls were consistently slower at the block-construction task than boys. In addition, older children's speed was predicted by their stereotypical attitudes towards LEGO blocks, whether their favorite toy was LEGO blocks, and how frequently they played with blocks at home. Results suggest that gender stereotypes affect children's performance on a gendered activity as early as age 4 and may be continuously affecting older girls' performance. Implications for understanding precursors of the STEM gender gap are discussed.