The paper introduces the R package betareg for beta regression, which is used to model variables that take values in the unit interval (0,1), such as rates and proportions. Beta regression assumes that the dependent variable follows a beta distribution, with its mean related to regressors through a link function. The model also includes a precision parameter, which can vary with regressors, allowing for heteroskedasticity and skewness. The package provides functions for estimation, inference, and diagnostics, and includes methods for fixed and variable dispersion beta regression. The paper describes the theoretical basis of beta regression, its implementation in R, and provides empirical examples using real data. It also discusses extensions of the model, such as variable dispersion beta regression, and highlights the flexibility of the beta distribution in modeling various types of data. The paper concludes with a discussion of future research directions, including the modeling of data with zeros and ones and the development of dynamic beta regression models.The paper introduces the R package betareg for beta regression, which is used to model variables that take values in the unit interval (0,1), such as rates and proportions. Beta regression assumes that the dependent variable follows a beta distribution, with its mean related to regressors through a link function. The model also includes a precision parameter, which can vary with regressors, allowing for heteroskedasticity and skewness. The package provides functions for estimation, inference, and diagnostics, and includes methods for fixed and variable dispersion beta regression. The paper describes the theoretical basis of beta regression, its implementation in R, and provides empirical examples using real data. It also discusses extensions of the model, such as variable dispersion beta regression, and highlights the flexibility of the beta distribution in modeling various types of data. The paper concludes with a discussion of future research directions, including the modeling of data with zeros and ones and the development of dynamic beta regression models.