diff --git a/markdownify/__init__.py b/markdownify/__init__.py
index 28cdaf6..94e18d7 100644
--- a/markdownify/__init__.py
+++ b/markdownify/__init__.py
@@ -41,6 +41,11 @@
 # confused with a list item
 re_escape_misc_list_items = re.compile(r'((?:\s|^)[0-9]{1,9})([.)](?:\s|$))')
 
+# Escape '|' characters in table cells, where they would otherwise be
+# parsed as column delimiters (even inside code spans). The negative
+# lookbehind skips pipes that escape_misc has already escaped.
+re_table_cell_pipe = re.compile(r'(?<!\\)\|')
+
 # Find consecutive backtick sequences in a string
 re_backtick_runs = re.compile(r'`+')
 
@@ -736,13 +741,15 @@ def convert_td(self, el, text, parent_tags):
         colspan = 1
         if 'colspan' in el.attrs and el['colspan'].isdigit():
             colspan = max(1, min(1000, int(el['colspan'])))
-        return ' ' + text.strip().replace("\n", " ") + ' |' * colspan
+        text = re_table_cell_pipe.sub(r'\\|', text.strip().replace("\n", " "))
+        return ' ' + text + ' |' * colspan
 
     def convert_th(self, el, text, parent_tags):
         colspan = 1
         if 'colspan' in el.attrs and el['colspan'].isdigit():
             colspan = max(1, min(1000, int(el['colspan'])))
-        return ' ' + text.strip().replace("\n", " ") + ' |' * colspan
+        text = re_table_cell_pipe.sub(r'\\|', text.strip().replace("\n", " "))
+        return ' ' + text + ' |' * colspan
 
     def convert_tr(self, el, text, parent_tags):
         cells = el.find_all(['td', 'th'])
diff --git a/tests/test_tables.py b/tests/test_tables.py
index 7e0670c..bef09c6 100644
--- a/tests/test_tables.py
+++ b/tests/test_tables.py
@@ -1,3 +1,5 @@
+import re
+
 from .utils import md
 
 
@@ -319,3 +321,24 @@ def test_table_infer_header():
     assert md(table_with_colspan, table_infer_header=True) == '\n\n| Name | | Age |\n| --- | --- | --- |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
     assert md(table_with_undefined_colspan, table_infer_header=True) == '\n\n| Name | Age |\n| --- | --- |\n| Jill | Smith |\n\n'
     assert md(table_with_colspan_missing_head, table_infer_header=True) == '\n\n| Name | | Age |\n| --- | --- | --- |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
+
+
+def test_pipes_in_cells():
+    # an unescaped '|' in cell content would be parsed as a column delimiter
+    assert md('<table><tr><th>A</th><th>B</th></tr><tr><td>1|2</td><td>3</td></tr></table>') == '\n\n| A | B |\n| --- | --- |\n| 1\\|2 | 3 |\n\n'
+    assert md('<table><tr><th>A|B</th><th>C</th></tr><tr><td>1</td><td>2</td></tr></table>') == '\n\n| A\\|B | C |\n| --- | --- |\n| 1 | 2 |\n\n'
+    assert md('<table><tr><th>A</th></tr><tr><td>1|2|3</td></tr></table>') == '\n\n| A |\n| --- |\n| 1\\|2\\|3 |\n\n'
+    assert md('<table><tr><th>A</th></tr><tr><td>|</td></tr></table>') == '\n\n| A |\n| --- |\n| \\| |\n\n'
+    assert md('<table><tr><th>A</th></tr><tr><td><b>1|2</b></td></tr></table>') == '\n\n| A |\n| --- |\n| **1\\|2** |\n\n'
+    assert md('<table><tr><th>A</th></tr><tr><td><a href="#">1|2</a></td></tr></table>') == '\n\n| A |\n| --- |\n| [1\\|2](#) |\n\n'
+    # GFM tables require pipes to be escaped even inside code spans
+    assert md('<table><tr><th>A</th></tr><tr><td><code>1|2</code></td></tr></table>') == '\n\n| A |\n| --- |\n| `1\\|2` |\n\n'
+    assert md('<table><tr><th>A</th><th>B</th></tr><tr><td colspan="2">1|2</td></tr></table>') == '\n\n| A | B |\n| --- | --- |\n| 1\\|2 | |\n\n'
+    # escape_misc=True escapes pipes on its own; make sure they are not escaped twice
+    assert md('<table><tr><th>A</th></tr><tr><td>1|2</td></tr></table>', escape_misc=True) == '\n\n| A |\n| --- |\n| 1\\|2 |\n\n'
+    # pipes outside of tables are unaffected
+    assert md('1|2') == '1|2'
+
+    # every row keeps the delimiter count of a two-column table
+    result = md('<table><tr><th>A|B</th><th>C</th></tr><tr><td>1|2</td><td>3|4</td></tr></table>')
+    assert [len(re.findall(r'(?<!\\)\|', line)) for line in result.strip().split('\n')] == [3, 3, 3]