There is a longstanding debate on the role of genetic ancestry in tuberculosis (TB) risk. While associations have been reported in the past, drawing convincing conclusions has been difficult since most of the previous studies could not adequately control for socioeconomic and environmental variables. Using genotyping data from 2,160 admixed Peruvians with active TB and 1,820 household contacts with latent TB, we inferred the level of Native American, European, African, and Asian ancestry per individual and investigated the relationship between TB progression risk and ancestry proportions. We observed a strong association between higher Native American ancestry proportion and increased risk of TB progression in the Peruvian population (p-value=4.2x10-15, OR=12.4, CI=11.8-13.1). This effect could not be explained by environmental or socioeconomic variables even after stringent permutation analysis within each household. We used admixture mapping to search for specific genomic regions that might explain some of this association. No genomic region reached the genome-wide significance threshold (p-value<5.7x10-5, lowest observed p-value=2x10-4). Regions with lower p-values were enriched for genes involved in T-cell (GO:0002456, enrichment p-values: 0.2, 0.02, and 0.007 for bins with p-value <0.01, <0.001, and <0.0005 respectively) and B-cell mediated immunity (GO:0019724, enrichment p-values: 0.2, 0.03, and 0.01) but not for the genes involved in innate immune responses (GO:0045087, enrichment p-values: 0.4, 0.2, and 0.1). Our results bring convincing evidence on the role of Native American ancestry in TB progression and suggest that this effect follows a polygenic architecture, including many genes that are involved in adaptive immune regulation of TB progression.