ADMIXTURE uses STRUCTURE's likelihood model for ancestry estimation. ADMIXTURE solves large problems in minutes while STRUCTURE takes hours. ADMIXTURE matches EIGENSTRAT's speed on many datasets. ADMIXTURE outperforms FRAPPE in speed and accuracy. ADMIXTURE estimates admixture coefficients and ancestral allele frequencies precisely. ADMIXTURE enables using vastly more markers in ancestry analysis. ADMIXTURE corrects population stratification in genetic studies effectively. ADMIXTURE employs block relaxation and quasi-Newton acceleration for speed. ADMIXTURE delivers maximum likelihood estimates comparable to STRUCTURE's Bayesian ones. ADMIXTURE works on unrelated individuals' multi-locus genotype data. ADMIXTURE handles datasets with 1,000 individuals and 10,000 markers swiftly.