The R2WinBUGS package allows users to run WinBUGS from within R. It automatically formats data and scripts for WinBUGS processing in batch mode. After WinBUGS completes, results can be read back into R for analysis, either through the package itself or using the coda package. Examples demonstrate the package's usage. WinBUGS is a software for Bayesian analysis using MCMC methods. It uses Gibbs sampling and the Metropolis algorithm to generate Markov chains. The R2WinBUGS package facilitates integration with R, enabling users to manipulate data in R, call WinBUGS, and analyze results. It supports batch processing and provides tools for convergence diagnostics and graphical summaries. The package includes functions to write data and scripts, run WinBUGS, and retrieve results. It also allows for posterior predictive simulations and graphical displays. The package is compatible with R and can be installed via CRAN. It is also possible to port R2WinBUGS to S-PLUS with minor modifications. The paper presents two examples: one involving educational testing data and another on childhood leukemia incidence related to benzene emissions. The first example uses a hierarchical normal model, while the second uses a Poisson-Gamma model. Both examples demonstrate the package's functionality in analyzing data and interpreting results. The R2WinBUGS package provides a wrapper function, bugs(), which handles data input, model execution, and result retrieval. It supports various arguments for controlling the MCMC process, including burn-in, iterations, thinning, and convergence diagnostics. The package also integrates with the coda package for further analysis of WinBUGS output. The package is useful for users who need to process data in R and analyze results from WinBUGS. It simplifies the workflow by allowing data manipulation in R, calling WinBUGS, and analyzing results within R. The package is particularly useful for handling large datasets and complex models. It also supports the use of the coda package for advanced analysis of MCMC output.The R2WinBUGS package allows users to run WinBUGS from within R. It automatically formats data and scripts for WinBUGS processing in batch mode. After WinBUGS completes, results can be read back into R for analysis, either through the package itself or using the coda package. Examples demonstrate the package's usage. WinBUGS is a software for Bayesian analysis using MCMC methods. It uses Gibbs sampling and the Metropolis algorithm to generate Markov chains. The R2WinBUGS package facilitates integration with R, enabling users to manipulate data in R, call WinBUGS, and analyze results. It supports batch processing and provides tools for convergence diagnostics and graphical summaries. The package includes functions to write data and scripts, run WinBUGS, and retrieve results. It also allows for posterior predictive simulations and graphical displays. The package is compatible with R and can be installed via CRAN. It is also possible to port R2WinBUGS to S-PLUS with minor modifications. The paper presents two examples: one involving educational testing data and another on childhood leukemia incidence related to benzene emissions. The first example uses a hierarchical normal model, while the second uses a Poisson-Gamma model. Both examples demonstrate the package's functionality in analyzing data and interpreting results. The R2WinBUGS package provides a wrapper function, bugs(), which handles data input, model execution, and result retrieval. It supports various arguments for controlling the MCMC process, including burn-in, iterations, thinning, and convergence diagnostics. The package also integrates with the coda package for further analysis of WinBUGS output. The package is useful for users who need to process data in R and analyze results from WinBUGS. It simplifies the workflow by allowing data manipulation in R, calling WinBUGS, and analyzing results within R. The package is particularly useful for handling large datasets and complex models. It also supports the use of the coda package for advanced analysis of MCMC output.