diff --git a/R-package/R/xgb.importance.R b/R-package/R/xgb.importance.R
index c1b45e81bb8c..be02b8cb0498 100644
--- a/R-package/R/xgb.importance.R
+++ b/R-package/R/xgb.importance.R
@@ -33,7 +33,9 @@
 #' For a linear model:
 #' - `Features`: Names of the features used in the model.
 #' - `Weight`: Linear coefficient of this feature.
-#' - `Class`: Class label (only for multiclass models).
+#' - `Class`: Class label (only for multiclass models). For objects of class `xgboost` (as
+#'   produced by [xgboost()]), it will be a `factor`, while for objects of class `xgb.Booster`
+#'   (as produced by [xgb.train()]), it will be a zero-based integer vector.
 #'
 #' If `feature_names` is not provided and `model` doesn't have `feature_names`,
 #' the index of the features will be used instead. Because the index is extracted from the model dump
@@ -144,11 +146,19 @@ xgb.importance <- function(model = NULL, feature_names = getinfo(model, "feature
         n_classes <- 0
     }
     importance <- if (n_classes == 0) {
-      data.table(Feature = results$features, Weight = results$weight)[order(-abs(Weight))]
+      return(data.table(Feature = results$features, Weight = results$weight)[order(-abs(Weight))])
     } else {
-      data.table(
+      out <- data.table(
         Feature = rep(results$features, each = n_classes), Weight = results$weight, Class = seq_len(n_classes) - 1
       )[order(Class, -abs(Weight))]
+      if (inherits(model, "xgboost") && NROW(attributes(model)$metadata$y_levels)) {
+        class_vec <- out$Class
+        class_vec <- as.integer(class_vec) + 1L
+        attributes(class_vec)$levels <- attributes(model)$metadata$y_levels
+        attributes(class_vec)$class <- "factor"
+        out[, Class := class_vec]
+      }
+      return(out[])
     }
   } else {
     concatenated <- list()
diff --git a/R-package/man/xgb.importance.Rd b/R-package/man/xgb.importance.Rd
index f26067d7fef9..503056737dd1 100644
--- a/R-package/man/xgb.importance.Rd
+++ b/R-package/man/xgb.importance.Rd
@@ -48,7 +48,9 @@ For a linear model:
 \itemize{
 \item \code{Features}: Names of the features used in the model.
 \item \code{Weight}: Linear coefficient of this feature.
-\item \code{Class}: Class label (only for multiclass models).
+\item \code{Class}: Class label (only for multiclass models). For objects of class \code{xgboost} (as
+produced by \code{\link[=xgboost]{xgboost()}}), it will be a \code{factor}, while for objects of class \code{xgb.Booster}
+(as produced by \code{\link[=xgb.train]{xgb.train()}}), it will be a zero-based integer vector.
 }
 
 If \code{feature_names} is not provided and \code{model} doesn't have \code{feature_names},
diff --git a/R-package/tests/testthat/test_xgboost.R b/R-package/tests/testthat/test_xgboost.R
index f3278cf37ea2..f7d9c367b8b9 100644
--- a/R-package/tests/testthat/test_xgboost.R
+++ b/R-package/tests/testthat/test_xgboost.R
@@ -1013,3 +1013,20 @@ test_that("'eval_set' as fraction works", {
   expect_true(hasName(evaluation_log, "eval_mlogloss"))
   expect_equal(length(attributes(model)$metadata$y_levels), 3L)
 })
+
+test_that("Linear booster importance uses class names", {
+  y <- iris$Species
+  x <- iris[, -5L]
+  model <- xgboost(
+    x,
+    y,
+    nthreads = 1L,
+    nrounds = 4L,
+    verbosity = 0L,
+    booster = "gblinear",
+    learning_rate = 0.2
+  )
+  imp <- xgb.importance(model)
+  expect_true(is.factor(imp$Class))
+  expect_equal(levels(imp$Class), levels(y))
+})