diff --git a/NAMESPACE b/NAMESPACE
index daa97ca16..93da52999 100644
--- a/NAMESPACE
+++ b/NAMESPACE
@@ -253,6 +253,7 @@ importFrom(utils,tail)
 importFrom(xml2,xml_attr)
 importFrom(xml2,xml_children)
 importFrom(xml2,xml_contents)
+importFrom(xml2,xml_find_all)
 importFrom(xml2,xml_name)
 importFrom(xml2,xml_text)
 importFrom(xml2,xml_type)
diff --git a/NEWS.md b/NEWS.md
index 6682a83ca..7b0f807b6 100644
--- a/NEWS.md
+++ b/NEWS.md
@@ -1,5 +1,8 @@
 # roxygen2 (development version)
 
+* Markdown tables with cells that contain multiple elements (e.g. text and code)
+  are now rendered correctly (#985).
+
 * `@includeRmd` has now an optional second argument, the top level section
   the included file will go to. It defaults to the details section (#970).
 
diff --git a/R/markdown.R b/R/markdown.R
index 1a943729c..45506e609 100644
--- a/R/markdown.R
+++ b/R/markdown.R
@@ -31,7 +31,7 @@ mdxml_children_to_rd <- function(xml, state) {
   paste0(out, collapse = "")
 }
 
-#' @importFrom xml2 xml_name xml_type xml_text xml_contents xml_attr xml_children
+#' @importFrom xml2 xml_name xml_type xml_text xml_contents xml_attr xml_children xml_find_all
 mdxml_node_to_rd <- function(xml, state) {
   if (!inherits(xml, "xml_node") ||
       ! xml_type(xml) %in% c("text", "element")) {
@@ -135,10 +135,11 @@ mdxml_table <- function(xml, state) {
   head <- xml_children(xml)[[1]]
   align <- substr(xml_attr(xml_children(head), "align", default = "left"), 1, 1)
 
-  rows <- xml_children(xml)
-  rows <- map(rows, xml_children)
-  rows_rd <- map(rows, ~ map_chr(xml_children(.x), mdxml_node_to_rd, state = state))
-  rows_rd <- map_chr(rows_rd, paste0, collapse = " \\tab ")
+  rows <- xml_find_all(xml, "d1:table_row|d1:table_header")
+  cells <- map(rows, xml_find_all, "d1:table_cell")
+
+  cells_rd <- map(cells, ~ map(.x, mdxml_children_to_rd, state = state))
+  rows_rd <- map_chr(cells_rd, paste0, collapse = " \\tab ")
 
   paste0("\\tabular{", paste(align, collapse = ""), "}{\n",
     paste("  ", rows_rd, "\\cr\n", collapse = ""),
diff --git a/tests/testthat/test-markdown-table.txt b/tests/testthat/test-markdown-table.txt
index 28d3ea51b..119ba9bff 100644
--- a/tests/testthat/test-markdown-table.txt
+++ b/tests/testthat/test-markdown-table.txt
@@ -1,11 +1,32 @@
-> cat(markdown(
-+   "\n| x   | y   |\n| --- | --- |\n| 1   | 2   |\n\n| x   | y   |\n| :-: | --: |\n| 1   | 2   |\n    "))
+> for (table in tables) {
++   cat_line(table)
++   cat_line(markdown(table))
++   cat_line()
++ }
+
+| x   | y   |
+| --- | --- |
+| 1   | 2   |
 \tabular{ll}{
    x \tab y \cr
    1 \tab 2 \cr
 }
+
+| x   | y   |
+| :-: | --: |
+| 1   | 2   |
 \tabular{cr}{
    x \tab y \cr
    1 \tab 2 \cr
 }
 
+| x     | y         |
+| ----- | --------- |
+| 1 _2_ | 3 *4* `5` |
+  
+\tabular{ll}{
+   x \tab y \cr
+   1 \emph{2} \tab 3 \emph{4} \code{5} \cr
+}
+
+
diff --git a/tests/testthat/test-markdown.R b/tests/testthat/test-markdown.R
index 02ea07de9..be1dd2aaf 100644
--- a/tests/testthat/test-markdown.R
+++ b/tests/testthat/test-markdown.R
@@ -228,17 +228,30 @@ test_that("nested lists are OK", {
 
 
 test_that("can convert table to Rd", {
+  txt <- "
+    | x   | y   |
+    | --- | --- |
+    | 1   | 2   |
+
+    | x   | y   |
+    | :-: | --: |
+    | 1   | 2   |
+
+    | x     | y         |
+    | ----- | --------- |
+    | 1 _2_ | 3 *4* `5` |
+  "
+  txt <- gsub("\n    ", "\n", txt)
+  tables <- strsplit(txt, "\n\n")[[1]]
+
   verify_output(
-    test_path("test-markdown-table.txt"),
-    cat(markdown("
-| x   | y   |
-| --- | --- |
-| 1   | 2   |
-
-| x   | y   |
-| :-: | --: |
-| 1   | 2   |
-    "))
+    test_path("test-markdown-table.txt"), {
+      for (table in tables) {
+        cat_line(table)
+        cat_line(markdown(table))
+        cat_line()
+      }
+    }
   )
 })