diff --git a/src/textual/widgets/_markdown.py b/src/textual/widgets/_markdown.py
index 5deb964d4a..9dd6f23699 100644
--- a/src/textual/widgets/_markdown.py
+++ b/src/textual/widgets/_markdown.py
@@ -294,10 +294,11 @@ def build_from_token(self, token: Token) -> None:
             token: The token from which this block is built.
         """
         self._inline_token = token
-        content = self._token_to_content(token)
+        content = self._markdown._get_token_content(token, block=self)
         self.set_content(content)
 
-    def _token_to_content(self, token: Token) -> Content:
+    @staticmethod
+    def _token_to_content(token: Token) -> Content:
         """Convert an inline token to Textual Content.
 
         Args:
@@ -1282,6 +1283,14 @@ async def load(self, path: Path | MarkdownLocation) -> None:
         """
         await self.sanitize_location(path).load(self)
 
+    def _get_token_content(self, token: Token, *, block: MarkdownBlock) -> Content:
+        """Supply inline content for every block, including headings and tables.
+
+        Prepared documents may return their acquired content here; ordinary
+        documents use the same native conversion without a separate parser.
+        """
+        return block._token_to_content(token)
+
     def unhandled_token(self, token: Token) -> MarkdownBlock | None:
         """Process an unhandled token.
 
diff --git a/tests/test_markdown.py b/tests/test_markdown.py
index fda3ba7d24..502f3d2098 100644
--- a/tests/test_markdown.py
+++ b/tests/test_markdown.py
@@ -12,6 +12,7 @@
 import textual.widgets._markdown as MD
 from textual import on
 from textual.app import App, ComposeResult
+from textual.content import Content
 from textual.style import Style
 from textual.widget import Widget
 from textual.widgets import Markdown
@@ -41,6 +42,35 @@ def compose(self) -> ComposeResult:
         yield FussyMarkdown(self._markdown)
 
 
+async def test_document_supplies_inline_content_for_heading_paragraph_and_table():
+    class CustomParagraph(MD.MarkdownParagraph):
+        def _token_to_content(self, token: Token) -> Content:
+            return super()._token_to_content(token) + Content(" custom")
+
+    class SuppliedMarkdown(Markdown):
+        BLOCKS = {**Markdown.BLOCKS, "paragraph_open": CustomParagraph}
+
+        def _get_token_content(self, token: Token, *, block: MarkdownBlock) -> Content:
+            return super()._get_token_content(token, block=block) + Content(" supplied")
+
+    document = SuppliedMarkdown(
+        "# Heading\n\nParagraph [link](https://example.com)\n\n"
+        "| Header |\n| --- |\n| Cell |"
+    )
+    async with App().run_test() as pilot:
+        await pilot.app.mount(document)
+        await pilot.pause()
+        assert document.query_one(MD.MarkdownH1)._content.plain == "Heading supplied"
+        paragraph = document.query_one(MD.MarkdownParagraph)._content
+        assert paragraph.plain == "Paragraph link custom supplied"
+        assert paragraph.spans == [
+            MD.Span(10, 14, Style.from_meta({"@click": "link('https://example.com')"}))
+        ]
+        table = document.query_one(MD.MarkdownTable)
+        assert [content.plain for content in table._headers] == ["Header supplied"]
+        assert [[content.plain for content in row] for row in table._rows] == [["Cell supplied"]]
+
+
 @pytest.mark.parametrize(
     ["document", "expected_nodes"],
     [