The pls package in R implements principal component regression (PCR) and partial least squares regression (PLSR). It is freely available from CRAN and licensed under the GNU General Public License. The package uses a formula interface similar to R's standard functions, making it easy for R users to interact with models. It includes methods for generic functions like predict, update, and coef, as well as specialized functions for scores, loadings, and RMSEP. The package also has a flexible cross-validation system and plot functions for visual inspection. The package implements PCR and several PLSR algorithms. It is modular, allowing easy integration into other functions. It is designed for interactive data analysis and as a building block for other packages using PLSR or PCR. The package includes a section on formulas and data frames for users unfamiliar with R's modeling idioms. PLSR and PCR are multivariate regression methods used in various fields, including chemistry. PLSR incorporates information from both X and Y, while PCR focuses on X. PLSR often requires fewer components than PCR to achieve similar prediction accuracy. Both methods behave as shrinkage methods, though PLSR may increase variance in regression coefficients. The paper describes the package and its use in data analysis, as well as its integration into other packages. It includes an example session using the gasoline data set to illustrate PLSR. The example shows how to fit a PLSR model, validate it, and predict new observations. The package also includes functions for formulas and data frames, model fitting, cross-validation, and model inspection. The package is designed to be flexible and efficient, with support for various algorithms and data handling. It includes functions for direct calls to fit functions, allowing users to customize the fitting process. The package is also used in other packages, such as lspls and plsgenomics, which combine PLSR with other methods. The package provides tools for model validation, prediction, and visualization, making it a valuable resource for multivariate regression analysis.The pls package in R implements principal component regression (PCR) and partial least squares regression (PLSR). It is freely available from CRAN and licensed under the GNU General Public License. The package uses a formula interface similar to R's standard functions, making it easy for R users to interact with models. It includes methods for generic functions like predict, update, and coef, as well as specialized functions for scores, loadings, and RMSEP. The package also has a flexible cross-validation system and plot functions for visual inspection. The package implements PCR and several PLSR algorithms. It is modular, allowing easy integration into other functions. It is designed for interactive data analysis and as a building block for other packages using PLSR or PCR. The package includes a section on formulas and data frames for users unfamiliar with R's modeling idioms. PLSR and PCR are multivariate regression methods used in various fields, including chemistry. PLSR incorporates information from both X and Y, while PCR focuses on X. PLSR often requires fewer components than PCR to achieve similar prediction accuracy. Both methods behave as shrinkage methods, though PLSR may increase variance in regression coefficients. The paper describes the package and its use in data analysis, as well as its integration into other packages. It includes an example session using the gasoline data set to illustrate PLSR. The example shows how to fit a PLSR model, validate it, and predict new observations. The package also includes functions for formulas and data frames, model fitting, cross-validation, and model inspection. The package is designed to be flexible and efficient, with support for various algorithms and data handling. It includes functions for direct calls to fit functions, allowing users to customize the fitting process. The package is also used in other packages, such as lspls and plsgenomics, which combine PLSR with other methods. The package provides tools for model validation, prediction, and visualization, making it a valuable resource for multivariate regression analysis.