Latin America continues to be severely underrepresented in genomics research, and fine-scale genetic histories as well as complex trait architectures remain hidden due to the lack of Big Data. To fill this gap, the Mexican Biobank project genotyped 1.8 million markers in 6,057 individuals from 32 states and 898 sampling localities across Mexico with linked complex trait and disease information creating a valuable nationwide genotype-phenotype database. Through a suite of state-of-the-art methods for ancestry deconvolution and inference of identity-by-descent (IBD) segments, we inferred detailed ancestral histories for the last 200 generations in different Mesoamerican regions, unravelling native and colonial/post-colonial demographic dynamics. We observed large variations in runs of homozygosity (ROH) among genomic regions with different ancestral origins reflecting their demographic histories, which also affect the distribution of rare deleterious variants across Mexico. We analysed a range of biomedical complex traits and identified significant genetic and environmental factors explaining their variation, such as ROH found to be significant predictors for trait variation in BMI and triglycerides. ====================================== This dataset contains GWAS summary statistics for the Mexico Biobank Project. Summary statistics for 22 binary and quantitative traits are provided from the full cohort of 5721 individuals from across Mexico, and a subset of 1061 individuals inferred to have more than 90% Native American ancestry.