This paper is a description of the algorithms and code that were developed in R to compute empirical Bayes estimates. The setting is that some variables that have direct survey estimates for some set of geographic areas or domains are given along with estimates of the variance of those variables. In addition, a set of variables that are correlated with the variable of interest is given. An empirical Bayes estimate is then a weighted average of the direct estimate and a regression estimate. It turns out that the empirical Bayes method in this setting is equivalent to a single level mixed model with known variances.