diff --git a/CHANGELOG.md b/CHANGELOG.md
index 2b090d416..37d0ebc38 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -4,6 +4,9 @@ Changelog
 NOTE: isort follows the [semver](https://semver.org/) versioning standard.
 Find out more about isort's release policy [here](https://pycqa.github.io/isort/docs/major_releases/release_policy/).
 
+### 5.6.3 October TBD, 2020
+  - Fixed #1556: Empty line added between imports that should be skipped.
+
 ### 5.6.2 October 10, 2020
   - Fixed #1548: On rare occasions an unecessary empty line can be added when an import is marked as skipped.
   - Fixed #1542: Bug in VERTICAL_PREFIX_FROM_MODULE_IMPORT wrap mode.
diff --git a/isort/parse.py b/isort/parse.py
index 96468f095..2c152651a 100644
--- a/isort/parse.py
+++ b/isort/parse.py
@@ -218,17 +218,35 @@ def file_contents(contents: str, config: Config = DEFAULT_CONFIG) -> ParsedConte
                 import_index = index - 1
                 while import_index and not in_lines[import_index - 1]:
                     import_index -= 1
-            elif "isort:skip" in line or "isort: skip" in line:
-                commentless = line.split("#", 1)[0]
+            else:
+                commentless = line.split("#", 1)[0].strip()
                 if (
-                    "(" in commentless
-                    and not commentless.rstrip().endswith(")")
-                    and import_index < line_count
+                    ("isort:skip" in line
+                    or "isort: skip" in line)
+                    and "(" in commentless
+                    and ")" not in commentless
                 ):
                     import_index = index
-                    while import_index < line_count and not commentless.rstrip().endswith(")"):
-                        commentless = in_lines[import_index].split("#", 1)[0]
-                        import_index += 1
+
+                    starting_line = line
+                    while "isort:skip" in starting_line or "isort: skip" in starting_line:
+                        commentless = starting_line.split("#", 1)[0]
+                        if (
+                            "(" in commentless
+                            and not commentless.rstrip().endswith(")")
+                            and import_index < line_count
+                        ):
+
+                            while import_index < line_count and not commentless.rstrip().endswith(")"):
+                                commentless = in_lines[import_index].split("#", 1)[0]
+                                import_index += 1
+                        else:
+                            import_index += 1
+
+                        if import_index >= line_count:
+                            break
+                        else:
+                            starting_line = in_lines[import_index]
 
         line, *end_of_line_comment = line.split("#", 1)
         if ";" in line:
diff --git a/tests/unit/test_regressions.py b/tests/unit/test_regressions.py
index 873cc2f72..c25ed19c7 100644
--- a/tests/unit/test_regressions.py
+++ b/tests/unit/test_regressions.py
@@ -1424,6 +1424,33 @@ def test_isort_shouldnt_split_skip_issue_1548():
     )
 
 
+def test_isort_shouldnt_split_skip_issue_1556():
+    assert isort.check_code(
+        """
+from tools.dependency_pruning.prune_dependencies import (  # isort:skip
+    prune_dependencies,
+)
+from tools.developer_pruning.prune_developers import (  # isort:skip
+    prune_developers,
+)
+""",
+        show_diff=True,
+        profile="black",
+        float_to_top=True,
+    )
+    assert isort.check_code(
+        """
+from tools.dependency_pruning.prune_dependencies import (  # isort:skip
+    prune_dependencies,
+)
+from tools.developer_pruning.prune_developers import x  # isort:skip
+""",
+        show_diff=True,
+        profile="black",
+        float_to_top=True,
+    )
+
+
 def test_isort_losing_imports_vertical_prefix_from_module_import_wrap_mode_issue_1542():
     """Ensure isort doesnt lose imports when a comment is combined with an import and
     wrap mode VERTICAL_PREFIX_FROM_MODULE_IMPORT is used.