Existing research on theories of intelligence shows that students with growth mindsets tend to outperform those with fixed mindsets in mathematics. We used nationally representative data to address two related questions in the general population: (a) Are there subgroup differences in the endorsement of a fixed mindset? (b) Does the negative association of a fixed mindset and math achievement vary across subgroups? We found that White students and students from higher socioeconomic backgrounds are more likely to view intelligence as a fixed trait than non-Whites and students from lower socioeconomic backgrounds. Furthermore, for low-achieving students, we found that a fixed mindset at 10th grade predicted lower gains in academic achievement by 12th grade than it did for their high-achieving counterparts. Our results reflect that contextual differences play a critical role in shaping fixed mindsets and its consequences.