The article introduces the BGLR R-package, which implements a variety of Bayesian regression models for genomic data analysis. These models include parametric variable selection and shrinkage methods, as well as semiparametric procedures such as Bayesian reproducing kernel Hilbert spaces regressions (RKHS). The package supports continuous, categorical, and censored response variables and is designed to handle large datasets with many predictors relative to the sample size. The authors describe the statistical models, algorithms, and data handling capabilities of BGLR, provide examples of its use with real and simulated data, and discuss practical issues in real-data analysis. The package is available on CRAN and R-forge, and includes two genomic datasets for illustration. The article also presents a benchmark evaluation of the computational performance of the package, highlighting its efficiency in handling large datasets.The article introduces the BGLR R-package, which implements a variety of Bayesian regression models for genomic data analysis. These models include parametric variable selection and shrinkage methods, as well as semiparametric procedures such as Bayesian reproducing kernel Hilbert spaces regressions (RKHS). The package supports continuous, categorical, and censored response variables and is designed to handle large datasets with many predictors relative to the sample size. The authors describe the statistical models, algorithms, and data handling capabilities of BGLR, provide examples of its use with real and simulated data, and discuss practical issues in real-data analysis. The package is available on CRAN and R-forge, and includes two genomic datasets for illustration. The article also presents a benchmark evaluation of the computational performance of the package, highlighting its efficiency in handling large datasets.