diff --git a/DESCRIPTION b/DESCRIPTION
index a2da2ea..fc2fa40 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -1,7 +1,7 @@
 Package: eiCompare
 Type: Package
 Title: Compares Different Ecological Inference Methods
-Version: 3.0.4
+Version: 3.0.5
 Authors@R: 
     c(person(given = "Loren",
              family = "Collingwood",
diff --git a/NAMESPACE b/NAMESPACE
index 8993f3c..c1483e2 100644
--- a/NAMESPACE
+++ b/NAMESPACE
@@ -15,6 +15,7 @@ export(ei_rc_good_table)
 export(ei_reg_bayes_conf_int)
 export(ei_rxc)
 export(elect_algebra)
+export(extract_rxc_precinct)
 export(fips_extract)
 export(get_multi_barreled_surnames)
 export(get_special_character_surnames)
diff --git a/NEWS.md b/NEWS.md
index 0019e75..e529c2b 100644
--- a/NEWS.md
+++ b/NEWS.md
@@ -1,7 +1,14 @@
+# eiCompare 3.0.5
+
+## added function 11/11/25
+
+* included extract_rxc_precinct() function to extract precinct level estimates from ei_rxc()
+
 # eiCompare 3.0.4
 
 ## Package changes
-
+* added rpv_normalize() function
+* removed wru dependency
 * incorporated rpv_coef_plot() and rpv_toDF() functions from eiExpand package
 * edited ei_iter() to have flexible CI parameters (default is 0.95) using bayestestR for calculation and updated column naming, and to use reproducible parallel processing (.inorder=TRUE)
 * edited ei_rxc() with repdocuible parallel processing and changed column naming to fit ei_iter()
diff --git a/R/extract_rxc_precinct.R b/R/extract_rxc_precinct.R
new file mode 100644
index 0000000..e1de4f1
--- /dev/null
+++ b/R/extract_rxc_precinct.R
@@ -0,0 +1,118 @@
+#' Extract Precinct-Level Estimates from ei.MD.bayes Object
+#'
+#' Extracts precinct-specific ecological inference estimates from ei_rxc() output.
+#' Uses exact string matching to handle variation column names
+#'
+#' @param eivote `ei_rxc()` output object containing `stat_objects`
+#' @param cand_cols Character vector of candidate column names (e.g., `c("pct_cand_A", "pct_cand_B")`)
+#' @param race_cols Character vector of race column names (e.g., `c("pct_black", "pct_white")`)
+#' @param dat Original data frame used in `ei_rxc()` call
+#' @param precinct_id Column name for precinct identifier (must exist in `dat`)
+#'
+#' @return Data frame with precinct IDs and race×candidate estimate columns
+#'
+#' @details
+#' The function extracts `md_out$draws$Beta` from the `ei_rxc()` output, which contains
+#' MCMC draws for each precinct-race-candidate combination. Beta column names follow
+#' the format `"beta.race_name.cand_name.precinct_idx"`. The function computes posterior
+#' means across MCMC iterations for each precinct.
+#'
+#' Output columns follow `expand.grid(cand, race)` ordering, with column names formatted
+#' as `paste0(race, cand)` (e.g., `"pct_blackpct_cand_A"`).
+#'
+#' @examples
+#' \donttest{
+#' 
+#' # library(eiCompare)
+#' # data(gwinnett_ei)
+#' #
+#' # gwinnett_ei$precinct <- 1:nrow(gwinnett_ei)
+#' #
+#' # eivote <- ei_rxc( #this will take some time
+#' #  data = gwinnett_ei,
+#' #  cand_cols = c("kemp", "abrams", "metz"),
+#' #  race_cols = c("white", "black", "other"),
+#' #  totals_col = "turnout",
+#' #  seed = 12345
+#' #)
+#'
+#' # # Extract precinct-level estimates
+#' # precinct_results <- extract_rxc_precinct(
+#' #  eivote = eivote,
+#' #  cand_cols = c("kemp", "abrams"),
+#' #  race_cols = c("white", "black", "other"),
+#' #  dat = gwinnett_ei,
+#' #  precinct_id = "precinct"
+#' #)
+#'
+#' #head(precinct_results)
+#' }
+#'
+#' @export
+extract_rxc_precinct <- function(eivote, cand_cols, race_cols, dat, precinct_id) {
+
+  # Extract md_out object from ei_rxc wrapper
+  eiMD_object <- eivote$stat_objects[[1]]
+
+  # Extract Beta matrix (MCMC iterations × beta parameters)
+  Beta <- eiMD_object$draws$Beta
+
+  # Check that precinct_id column exists in dat
+  if(!precinct_id %in% colnames(dat)) {
+    stop(paste0("Column '", precinct_id, "' not found in dat. ",
+                "Available columns: ", paste(colnames(dat), collapse = ", ")))
+  }
+
+  n_precincts <- nrow(dat)
+  beta_colnames <- colnames(Beta)
+
+  # Initialize result matrix (precincts × race-candidate combinations)
+  result_matrix <- matrix(NA,
+                          nrow = n_precincts,
+                          ncol = length(race_cols) * length(cand_cols))
+
+  # Loop through race-candidate combinations and extract precinct estimates
+  col_idx <- 1
+  for(race in race_cols) {
+    for(cand in cand_cols) {
+
+      # Build expected prefix pattern for exact matching
+      # Format: beta.race.cand.precinct_number
+      expected_prefix <- paste0("beta.", race, ".", cand, ".")
+
+      # Find Beta columns matching this race-candidate pair
+      matching_cols <- grep(paste0("^", gsub("\\.", "\\\\.", expected_prefix)),
+                           beta_colnames,
+                           value = FALSE)
+
+      # Validation - should have exactly n_precincts matches
+      if(length(matching_cols) != n_precincts) {
+        stop(paste0("Column matching error for race='", race, "', cand='", cand,
+                    "': found ", length(matching_cols), " columns but expected ",
+                    n_precincts, " precincts"))
+      }
+
+      # Extract precinct indices and reorder to match dat row order
+      precinct_nums <- sub(expected_prefix, "", beta_colnames[matching_cols])
+      precinct_order <- order(as.numeric(precinct_nums))
+      matching_cols_ordered <- matching_cols[precinct_order]
+
+      # Calculate mean across MCMC iterations for each precinct
+      result_matrix[, col_idx] <- colMeans(Beta[, matching_cols_ordered])
+      col_idx <- col_idx + 1
+    }
+  }
+
+  # Create column names (race + candidate, matching expand.grid order)
+  col_names_df <- expand.grid(cand = cand_cols, race = race_cols)
+  col_names <- paste0(col_names_df$race, col_names_df$cand)
+
+  # Convert to data frame with column names
+  result_df <- as.data.frame(result_matrix)
+  colnames(result_df) <- col_names
+
+  # Attach precinct IDs from original data as first column
+  result_df <- cbind(dat[, precinct_id, drop = FALSE], result_df)
+
+  return(result_df)
+}
diff --git a/README.md b/README.md
index 6353efe..1936cfe 100644
--- a/README.md
+++ b/README.md
@@ -13,6 +13,27 @@
 
 ## News
 
+# eiCompare 3.0.5
+
+## New function
+
+* included extract_rxc_precinct() function to extract precinct level estimates from ei_rxc()
+
+# eiCompare 3.0.4
+
+## Package changes
+* added add_rpv_normalize() function
+* removed wru dependency
+* incorporated rpv_coef_plot() and rpv_toDF() functions from eiExpand package
+* edited ei_iter() to have flexible CI parameters (default is 0.95) using bayestestR for calculation and updated column naming, and to use reproducible parallel processing (.inorder=TRUE)
+* edited ei_rxc() with repdocuible parallel processing and changed column naming to fit ei_iter()
+* Fixed summary.eiCompare() print behavior
+* Added viridis to imports for color visualiztion and updated RoxygenNote to 7.3.2
+
+### eiCompare 3.0.3
+
+Updated 
+
 ### eiCompare 3.0.2 
 
 #### Package changes
diff --git a/man/extract_rxc_precinct.Rd b/man/extract_rxc_precinct.Rd
new file mode 100644
index 0000000..b77b3cf
--- /dev/null
+++ b/man/extract_rxc_precinct.Rd
@@ -0,0 +1,64 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/extract_rxc_precinct.R
+\name{extract_rxc_precinct}
+\alias{extract_rxc_precinct}
+\title{Extract Precinct-Level Estimates from ei.MD.bayes Object}
+\usage{
+extract_rxc_precinct(eivote, cand_cols, race_cols, dat, precinct_id)
+}
+\arguments{
+\item{eivote}{`ei_rxc()` output object containing `stat_objects`}
+
+\item{cand_cols}{Character vector of candidate column names (e.g., `c("pct_cand_A", "pct_cand_B")`)}
+
+\item{race_cols}{Character vector of race column names (e.g., `c("pct_black", "pct_white")`)}
+
+\item{dat}{Original data frame used in `ei_rxc()` call}
+
+\item{precinct_id}{Column name for precinct identifier (must exist in `dat`)}
+}
+\value{
+Data frame with precinct IDs and race×candidate estimate columns
+}
+\description{
+Extracts precinct-specific ecological inference estimates from ei_rxc() output.
+Uses exact string matching to handle variation column names
+}
+\details{
+The function extracts `md_out$draws$Beta` from the `ei_rxc()` output, which contains
+MCMC draws for each precinct-race-candidate combination. Beta column names follow
+the format `"beta.race_name.cand_name.precinct_idx"`. The function computes posterior
+means across MCMC iterations for each precinct.
+
+Output columns follow `expand.grid(cand, race)` ordering, with column names formatted
+as `paste0(race, cand)` (e.g., `"pct_blackpct_cand_A"`).
+}
+\examples{
+\donttest{
+
+# library(eiCompare)
+# data(gwinnett_ei)
+#
+# gwinnett_ei$precinct <- 1:nrow(gwinnett_ei)
+#
+# eivote <- ei_rxc( #this will take some time
+#  data = gwinnett_ei,
+#  cand_cols = c("kemp", "abrams", "metz"),
+#  race_cols = c("white", "black", "other"),
+#  totals_col = "turnout",
+#  seed = 12345
+#)
+
+# # Extract precinct-level estimates
+# precinct_results <- extract_rxc_precinct(
+#  eivote = eivote,
+#  cand_cols = c("kemp", "abrams"),
+#  race_cols = c("white", "black", "other"),
+#  dat = gwinnett_ei,
+#  precinct_id = "precinct"
+#)
+
+#head(precinct_results)
+}
+
+}