diff --git a/pandas/io/formats/info.py b/pandas/io/formats/info.py
index e8e41d4325103..9812ca39e2656 100644
--- a/pandas/io/formats/info.py
+++ b/pandas/io/formats/info.py
@@ -288,7 +288,11 @@ def _verbose_repr(
         len_column = len(pprint_thing(column_head))
         space = max(max_col, len_column) + col_space
 
-        max_id = len(pprint_thing(col_count))
+        # GH #36765
+        # add one space in max_id because there is a one-space padding
+        # in front of the number
+        # this allows maintain two spaces gap between columns
+        max_id = len(pprint_thing(col_count)) + 1
         len_id = len(pprint_thing(id_head))
         space_num = max(max_id, len_id) + col_space
 
diff --git a/pandas/tests/io/formats/test_info.py b/pandas/tests/io/formats/test_info.py
index 7000daeb9b575..d98530b5435e7 100644
--- a/pandas/tests/io/formats/test_info.py
+++ b/pandas/tests/io/formats/test_info.py
@@ -87,7 +87,7 @@ def test_info_verbose():
     frame.info(verbose=True, buf=buf)
 
     res = buf.getvalue()
-    header = " #    Column  Dtype  \n---   ------  -----  "
+    header = " #     Column  Dtype  \n---    ------  -----  "
     assert header in res
 
     frame.info(verbose=True, buf=buf)
@@ -101,6 +101,64 @@ def test_info_verbose():
             assert line.startswith(line_nr)
 
 
+@pytest.mark.parametrize(
+    "size, header_exp, separator_exp, first_line_exp, last_line_exp",
+    [
+        (
+            4,
+            " #   Column  Non-Null Count  Dtype  ",
+            "---  ------  --------------  -----  ",
+            " 0   0       3 non-null      float64",
+            " 3   3       3 non-null      float64",
+        ),
+        (
+            11,
+            " #   Column  Non-Null Count  Dtype  ",
+            "---  ------  --------------  -----  ",
+            " 0   0       3 non-null      float64",
+            " 10  10      3 non-null      float64",
+        ),
+        (
+            101,
+            " #    Column  Non-Null Count  Dtype  ",
+            "---   ------  --------------  -----  ",
+            " 0    0       3 non-null      float64",
+            " 100  100     3 non-null      float64",
+        ),
+        (
+            1001,
+            " #     Column  Non-Null Count  Dtype  ",
+            "---    ------  --------------  -----  ",
+            " 0     0       3 non-null      float64",
+            " 1000  1000    3 non-null      float64",
+        ),
+        (
+            10001,
+            " #      Column  Non-Null Count  Dtype  ",
+            "---     ------  --------------  -----  ",
+            " 0      0       3 non-null      float64",
+            " 10000  10000   3 non-null      float64",
+        ),
+    ],
+)
+def test_info_verbose_with_counts_spacing(
+    size, header_exp, separator_exp, first_line_exp, last_line_exp
+):
+    """Test header column, spacer, first line and last line in verbose mode."""
+    frame = DataFrame(np.random.randn(3, size))
+    buf = StringIO()
+    frame.info(verbose=True, null_counts=True, buf=buf)
+    all_lines = buf.getvalue().splitlines()
+    # Here table would contain only header, separator and table lines
+    # dframe repr, index summary, memory usage and dtypes are excluded
+    table = all_lines[3:-2]
+    header, separator, first_line, *rest, last_line = table
+    assert header == header_exp
+    assert separator == separator_exp
+    assert first_line == first_line_exp
+    assert last_line == last_line_exp
+
+
 def test_info_memory():
     # https://github.com/pandas-dev/pandas/issues/21056
     df = DataFrame({"a": Series([1, 2], dtype="i8")})