diff --git a/novelwriter/core/index.py b/novelwriter/core/index.py
index c1cb1e01..cdd78032 100644
--- a/novelwriter/core/index.py
+++ b/novelwriter/core/index.py
@@ -7,6 +7,7 @@ File History:
 Created: 2019-04-22 [0.0.1]  countWords
 Created: 2019-05-27 [0.1.4]  NWIndex
 Created: 2022-05-28 [1.7rc1] IndexItem, IndexHeading
+Created: 2022-05-29 [1.7rc1] TagsIndex, ItemIndex
 
 This file is a part of novelWriter
 Copyright 2018–2022, Veronica Berglyd Olsen
@@ -43,10 +44,24 @@ logger = logging.getLogger(__name__)
 
 H_VALID = ("H0", "H1", "H2", "H3", "H4")
 H_LEVEL = {"H0": 0, "H1": 1, "H2": 2, "H3": 3, "H4": 4}
-H_NONE = "T000000"
+TT_NONE = "T000000"
 
 
 class NWIndex:
+    """This class holds the entire index for a given project. The index
+    contains the data that isn't stored in the project items themselves.
+    The content of the index is updated every time a file item is saved.
+
+    The primary index data is contained in the ItemIndex class, which
+    contains an IndexItem representing each NWItem. Each IndexItem holds
+    an IndexHeading object for each heading of the item's text.
+
+    A reverse index of all tags is contained in the TagsIndex class.
+    This is duplicate information used for quicker lookups from the tags
+    and back to items where they are defined.
+
+    The index data is cached in a JSON file between writing sessions.
+    """
 
     def __init__(self, theProject):
 
@@ -104,10 +119,10 @@ class NWIndex:
         moved from the archive or trash folders back into the active
         project.
         """
-        logger.debug("Re-indexing item '%s'", tHandle)
         if not self.theProject.tree.checkType(tHandle, nwItemType.FILE):
             return False
 
+        logger.debug("Re-indexing item '%s'", tHandle)
         theDoc = NWDoc(self.theProject, tHandle)
         theText = theDoc.readDocument()
         self.scanText(tHandle, theText if theText is not None else "")
@@ -140,6 +155,7 @@ class NWIndex:
         indexFile = os.path.join(self.theProject.projMeta, nwFiles.INDEX_FILE)
         tStart = time()
 
+        self._indexBroken = False
         if os.path.isfile(indexFile):
             logger.debug("Loading index file")
             try:
@@ -222,8 +238,9 @@ class NWIndex:
             logger.info("Not indexing non-file item '%s'", tHandle)
             return False
 
-        # Delete the old entry and create a new
-        self.deleteHandle(tHandle)
+        # Delete tags and create new item entry
+        for tTag in self._itemIndex.allItemTags(tHandle):
+            del self._tagsIndex[tTag]
         self._itemIndex.add(tHandle, theItem)
 
         # Run word counter for the whole text
@@ -644,7 +661,7 @@ class TagsIndex:
         """
         if tagKey in self._tags:
             return self._tags.get(tagKey).get("heading")
-        return H_NONE
+        return TT_NONE
 
     def tagClass(self, tagKey):
         """Get the class of a given tag.
@@ -782,11 +799,11 @@ class ItemIndex:
                 continue
 
             if rootHandle is None:
-                for sTitle, hItem in self._items[tHandle].items():
-                    yield tHandle, sTitle, hItem
-            elif tItem.rootHandle == rootHandle:
-                for sTitle, hItem in self._items[tHandle].items():
-                    yield tHandle, sTitle, hItem
+                for sTitle in self._items[tHandle].headings():
+                    yield tHandle, sTitle, self._items[tHandle][sTitle]
+            elif tItem.itemRoot == rootHandle:
+                for sTitle in self._items[tHandle].headings():
+                    yield tHandle, sTitle, self._items[tHandle][sTitle]
             else:
                 continue
 
@@ -876,7 +893,7 @@ class IndexItem:
         self._index = 0
 
         # Add a placeholder heading
-        self._headings[H_NONE] = IndexHeading(H_NONE)
+        self._headings[TT_NONE] = IndexHeading(TT_NONE)
 
         return
 
@@ -910,8 +927,8 @@ class IndexItem:
         """Add a heading to the item. Also remove the placeholder entry
         if it exists.
         """
-        if H_NONE in self._headings:
-            self._headings.pop(H_NONE)
+        if TT_NONE in self._headings:
+            self._headings.pop(TT_NONE)
         self._headings[tHeading.key] = tHeading
         return
 
@@ -957,6 +974,9 @@ class IndexItem:
     def items(self):
         return self._headings.items()
 
+    def headings(self):
+        return sorted(self._headings.keys())
+
     def allTags(self):
         """Return a list of all tags in the current item.
         """
@@ -1102,9 +1122,10 @@ class IndexHeading:
         """Add a record of a reference tag, and what keyword types it is
         associated with.
         """
-        if tagKey not in self._refs:
-            self._refs[tagKey] = set()
-        self._refs[tagKey].add(refType)
+        if refType in nwKeyWords.VALID_KEYS:
+            if tagKey not in self._refs:
+                self._refs[tagKey] = set()
+            self._refs[tagKey].add(refType)
         return
 
     ##
diff --git a/novelwriter/guimain.py b/novelwriter/guimain.py
index 3a957577..3d61a65e 100644
--- a/novelwriter/guimain.py
+++ b/novelwriter/guimain.py
@@ -865,22 +865,13 @@ class GuiMain(QMainWindow):
         self.theProject.index.clearIndex()
 
         for tItem in self.theProject.tree:
+            if tItem is None:  # pragma: no cover
+                continue  # This is a bug trap
 
-            if tItem is not None:
-                self.setStatus(self.tr("Indexing: '{0}'").format(tItem.itemName))
-            else:
-                self.setStatus(self.tr("Indexing: '{0}'").format(self.tr("Unknown item")))
-
-            if tItem is not None and tItem.itemType == nwItemType.FILE:
-                logger.verbose("Scanning '%s'", tItem.itemName)
-                self.theProject.index.reIndexHandle(tItem.itemHandle)
-
-                # Get Word Counts
-                cC, wC, pC = self.theProject.index.getCounts(tItem.itemHandle)
-                tItem.setCharCount(cC)
-                tItem.setWordCount(wC)
-                tItem.setParaCount(pC)
-                self.treeView.propagateCount(tItem.itemHandle, wC, countChildren=True)
+            logger.verbose("Indexing '%s'", tItem.itemName)
+            if self.theProject.index.reIndexHandle(tItem.itemHandle):
+                # Update Word Counts
+                self.treeView.propagateCount(tItem.itemHandle, tItem.wordCount, countChildren=True)
                 self.treeView.setTreeItemValues(tItem.itemHandle)
 
         tEnd = time()
diff --git a/sample/content/5eaea4e8cdee8.nwd b/sample/content/5eaea4e8cdee8.nwd
index 1a7f3c79..0f8ecc26 100644
--- a/sample/content/5eaea4e8cdee8.nwd
+++ b/sample/content/5eaea4e8cdee8.nwd
@@ -4,5 +4,6 @@
 # Mars
 
 @tag: Mars
+@location: Space
 
 It’s red. Dusty and red.
diff --git a/sample/content/88706ddc78b1b.nwd b/sample/content/88706ddc78b1b.nwd
index ce4d2123..0f140538 100644
--- a/sample/content/88706ddc78b1b.nwd
+++ b/sample/content/88706ddc78b1b.nwd
@@ -1,5 +1,5 @@
 %%~name: Chapter Two
-%%~path: e7ded148d6e4a/88706ddc78b1b
+%%~path: 7031beac91f75/88706ddc78b1b
 %%~kind: NOVEL/DOCUMENT
 ## Where has John Gone?
 
diff --git a/sample/content/b3e74dbc1f584.nwd b/sample/content/b3e74dbc1f584.nwd
index 6931a299..c980865e 100644
--- a/sample/content/b3e74dbc1f584.nwd
+++ b/sample/content/b3e74dbc1f584.nwd
@@ -4,5 +4,6 @@
 # Earth
 
 @tag: Earth
+@location: Space
 
 Third planet from the sun, fairly dense, and with lots of people on it.
\ No newline at end of file
diff --git a/sample/content/b8136a5a774a0.nwd b/sample/content/b8136a5a774a0.nwd
index 3c2c1854..636c7227 100644
--- a/sample/content/b8136a5a774a0.nwd
+++ b/sample/content/b8136a5a774a0.nwd
@@ -1,6 +1,6 @@
 %%~name: Delete Me!
 %%~path: 98acd8c76c93a/b8136a5a774a0
-%%~kind: NOVEL/DOCUMENT
+%%~kind: TRASH/DOCUMENT
 ### Delete Me!
 
 This scene is trash.
\ No newline at end of file
diff --git a/tests/test_core/test_core_index.py b/tests/test_core/test_core_index.py
index 76ffc70a..b1233275 100644
--- a/tests/test_core/test_core_index.py
+++ b/tests/test_core/test_core_index.py
@@ -19,14 +19,14 @@ You should have received a copy of the GNU General Public License
 along with this program. If not, see <https://www.gnu.org/licenses/>.
 """
 
-import pytest
 import os
 import json
+import pytest
 
 from shutil import copyfile
 
 from mock import causeException
-from tools import cmpFiles
+from tools import buildTestProject, cmpFiles, writeFile
 
 from novelwriter.core.project import NWProject
 from novelwriter.core.index import NWIndex, countWords
@@ -87,36 +87,58 @@ def testCoreIndex_LoadSave(monkeypatch, nwLipsum, mockGUI, outDir, refDir):
     with monkeypatch.context() as mp:
         mp.setattr(json, "load", causeException)
         assert theIndex.loadIndex() is False
+        assert theIndex.indexBroken is True
 
     # Make the load pass
     assert theIndex.loadIndex() is True
+    assert theIndex.indexBroken is False
 
     assert str(theIndex._tagsIndex.packData()) == tagIndex
     assert str(theIndex._itemIndex.packData()) == itemsIndex
 
-    # Break the index and check that we notice
-    # assert theIndex.indexBroken is False
-    # theIndex._tagIndex["Bod"].append("Stuff")
-    # theIndex._checkIndex()
-    # assert theIndex.indexBroken is True
+    # Check File
+    copyfile(projFile, testFile)
+    assert cmpFiles(testFile, compFile)
+
+    # Write an emtpy index file and load it
+    writeFile(projFile, "{}")
+    assert theIndex.loadIndex() is False
+    assert theIndex.indexBroken is True
+
+    # Write an index file that passes loading, but is still empty
+    writeFile(projFile, '{"tagsIndex": {}, "itemIndex": {}}')
+    assert theIndex.loadIndex() is True
+    assert theIndex.indexBroken is False
+
+    # Check that the index is re-populated
+    assert "04468803b92e1" in theIndex._itemIndex
+    assert "2426c6f0ca922" in theIndex._itemIndex
+    assert "441420a886d82" in theIndex._itemIndex
+    assert "47666c91c7ccf" in theIndex._itemIndex
+    assert "4c4f28287af27" in theIndex._itemIndex
+    assert "846352075de7d" in theIndex._itemIndex
+    assert "88243afbe5ed8" in theIndex._itemIndex
+    assert "88d59a277361b" in theIndex._itemIndex
+    assert "8c58a65414c23" in theIndex._itemIndex
+    assert "db7e733775d4d" in theIndex._itemIndex
+    assert "eb103bc70c90c" in theIndex._itemIndex
+    assert "f8c0562e50f1b" in theIndex._itemIndex
+    assert "f96ec11c6a3da" in theIndex._itemIndex
+    assert "fb609cd8319dc" in theIndex._itemIndex
+    assert "7a992350f3eb6" in theIndex._itemIndex
 
     # Finalise
     assert theProject.closeProject() is True
 
-    copyfile(projFile, testFile)
-    assert cmpFiles(testFile, compFile)
-
 # END Test testCoreIndex_LoadSave
 
 
 @pytest.mark.core
-def testCoreIndex_ScanThis(nwMinimal, mockGUI):
+def testCoreIndex_ScanThis(mockGUI):
     """Test the tag scanner function scanThis.
     """
     theProject = NWProject(mockGUI)
-    assert theProject.openProject(nwMinimal) is True
-
-    theIndex = NWIndex(theProject)
+    theIndex = theProject.index
 
     isValid, theBits, thePos = theIndex.scanThis("tag: this, and this")
     assert isValid is False
@@ -161,15 +183,15 @@ def testCoreIndex_ScanThis(nwMinimal, mockGUI):
 
 
 @pytest.mark.core
-def testCoreIndex_CheckThese(nwMinimal, mockGUI):
+def testCoreIndex_CheckThese(mockGUI, fncDir, mockRnd):
     """Test the tag checker function checkThese.
     """
     theProject = NWProject(mockGUI)
-    assert theProject.openProject(nwMinimal) is True
+    buildTestProject(theProject, fncDir)
+    theIndex = theProject.index
 
-    theIndex = NWIndex(theProject)
-    nHandle = theProject.newFile("Hello", "a508bb932959c")
-    cHandle = theProject.newFile("Jane",  "afb3043c7b2b3")
+    nHandle = theProject.newFile("Hello", "0000000000010")
+    cHandle = theProject.newFile("Jane",  "0000000000012")
     nItem = theProject.tree[nHandle]
     cItem = theProject.tree[cHandle]
 
@@ -239,17 +261,16 @@ def testCoreIndex_CheckThese(nwMinimal, mockGUI):
 
 
 @pytest.mark.core
-def testCoreIndex_ScanText(nwMinimal, mockGUI):
+def testCoreIndex_ScanText(mockGUI, fncDir, mockRnd):
     """Check the index text scanner.
     """
     theProject = NWProject(mockGUI)
-    assert theProject.openProject(nwMinimal) is True
-
-    theIndex = NWIndex(theProject)
+    buildTestProject(theProject, fncDir)
+    theIndex = theProject.index
 
     # Some items for fail to scan tests
-    dHandle = theProject.newFolder("Folder", "a508bb932959c")
-    xHandle = theProject.newFile("No Layout", "a508bb932959c")
+    dHandle = theProject.newFolder("Folder", "0000000000010")
+    xHandle = theProject.newFile("No Layout", "0000000000010")
     xItem = theProject.tree[xHandle]
     xItem.setLayout(nwItemLayout.NO_LAYOUT)
 
@@ -279,11 +300,11 @@ def testCoreIndex_ScanText(nwMinimal, mockGUI):
     assert theIndex.scanText(xHandle, "Hello World!") is False
 
     # Make some usable items
-    tHandle = theProject.newFile("Title", "a508bb932959c")
-    pHandle = theProject.newFile("Page",  "a508bb932959c")
-    nHandle = theProject.newFile("Hello", "a508bb932959c")
-    cHandle = theProject.newFile("Jane",  "afb3043c7b2b3")
-    sHandle = theProject.newFile("Scene", "a508bb932959c")
+    tHandle = theProject.newFile("Title", "0000000000010")
+    pHandle = theProject.newFile("Page",  "0000000000010")
+    nHandle = theProject.newFile("Hello", "0000000000010")
+    cHandle = theProject.newFile("Jane",  "0000000000012")
+    sHandle = theProject.newFile("Scene", "0000000000010")
 
     # Text Indexing
     # =============
@@ -449,18 +470,27 @@ def testCoreIndex_ScanText(nwMinimal, mockGUI):
 
 
 @pytest.mark.core
-def testCoreIndex_ExtractData(nwMinimal, mockGUI):
+def testCoreIndex_ExtractData(mockGUI, fncDir, mockRnd):
     """Check the index data extraction functions.
     """
     theProject = NWProject(mockGUI)
-    assert theProject.openProject(nwMinimal) is True
+    buildTestProject(theProject, fncDir)
 
-    theIndex = NWIndex(theProject)
-    nHandle = theProject.newFile("Hello", "a508bb932959c")
-    cHandle = theProject.newFile("Jane",  "afb3043c7b2b3")
+    theIndex = theProject.index
+    theIndex.reIndexHandle("0000000000010")
+    theIndex.reIndexHandle("0000000000011")
+    theIndex.reIndexHandle("0000000000012")
+    theIndex.reIndexHandle("0000000000013")
+    theIndex.reIndexHandle("0000000000014")
+    theIndex.reIndexHandle("0000000000015")
+    theIndex.reIndexHandle("0000000000016")
+    theIndex.reIndexHandle("0000000000017")
+
+    nHandle = theProject.newFile("Hello", "0000000000010")
+    cHandle = theProject.newFile("Jane",  "0000000000012")
 
     assert theIndex.getNovelData("", "") is None
-    assert theIndex.getNovelData("a508bb932959c", "") is None
+    assert theIndex.getNovelData("0000000000010", "") is None
 
     assert theIndex.scanText(cHandle, (
         "# Jane Smith\n"
@@ -480,7 +510,12 @@ def testCoreIndex_ExtractData(nwMinimal, mockGUI):
     for aKey, _, _, _ in theIndex.novelStructure():
         theKeys.append(aKey)
 
-    assert theKeys == ["%s:T000001" % nHandle]
+    assert theKeys == [
+        "0000000000014:T000001",
+        "0000000000016:T000001",
+        "0000000000017:T000001",
+        "%s:T000001" % nHandle,
+    ]
 
     # Check that excluded files can be skipped
     theProject.tree[nHandle].setExported(False)
@@ -489,19 +524,22 @@ def testCoreIndex_ExtractData(nwMinimal, mockGUI):
     for aKey, _, _, _ in theIndex.novelStructure(skipExcl=False):
         theKeys.append(aKey)
 
-    assert theKeys == ["%s:T000001" % nHandle]
+    assert theKeys == [
+        "0000000000014:T000001",
+        "0000000000016:T000001",
+        "0000000000017:T000001",
+        "%s:T000001" % nHandle,
+    ]
 
     theKeys = []
     for aKey, _, _, _ in theIndex.novelStructure(skipExcl=True):
         theKeys.append(aKey)
 
-    assert theKeys == []
-
-    theKeys = []
-    for aKey, _, _, _ in theIndex.novelStructure():
-        theKeys.append(aKey)
-
-    assert theKeys == []
+    assert theKeys == [
+        "0000000000014:T000001",
+        "0000000000016:T000001",
+        "0000000000017:T000001",
+    ]
 
     # The novel file should have the correct counts
     cC, wC, pC = theIndex.getCounts(nHandle)
@@ -528,6 +566,9 @@ def testCoreIndex_ExtractData(nwMinimal, mockGUI):
     # None handle should return an empty dict
     assert theIndex.getBackReferenceList(None) == {}
 
+    # The Title Page file should have no references as it has no tag
+    assert theIndex.getBackReferenceList("0000000000014") == {}
+
     # The character file should have a record of the reference from the novel file
     theRefs = theIndex.getBackReferenceList(cHandle)
     assert theRefs == {nHandle: "T000001"}
@@ -542,6 +583,10 @@ def testCoreIndex_ExtractData(nwMinimal, mockGUI):
     # =========
     # For whole text and sections
 
+    # Invalid handle or title should return 0s
+    assert theIndex.getCounts("stuff") == (0, 0, 0)
+    assert theIndex.getCounts(nHandle, "stuff") == (0, 0, 0)
+
     # Get section counts for a novel file
     assert theIndex.scanText(nHandle, (
         "# Hello World!\n"
@@ -611,9 +656,9 @@ def testCoreIndex_ExtractData(nwMinimal, mockGUI):
     # Novel Stats
     # ===========
 
-    hHandle = theProject.newFile("Chapter", "a508bb932959c")
-    sHandle = theProject.newFile("Scene One", "a508bb932959c")
-    tHandle = theProject.newFile("Scene Two", "a508bb932959c")
+    hHandle = theProject.newFile("Chapter", "0000000000010")
+    sHandle = theProject.newFile("Scene One", "0000000000010")
+    tHandle = theProject.newFile("Scene Two", "0000000000010")
 
     theProject.tree[hHandle].itemLayout == nwItemLayout.DOCUMENT
     theProject.tree[sHandle].itemLayout == nwItemLayout.DOCUMENT
@@ -624,6 +669,9 @@ def testCoreIndex_ExtractData(nwMinimal, mockGUI):
     assert theIndex.scanText(tHandle, "### Scene Two\n\n")
 
     assert [(h, t) for h, t, _ in theIndex._itemIndex.iterNovelStructure(skipExcl=False)] == [
+        ("0000000000014", "T000001"),
+        ("0000000000016", "T000001"),
+        ("0000000000017", "T000001"),
         (nHandle, "T000001"),
         (nHandle, "T000011"),
         (hHandle, "T000001"),
@@ -632,6 +680,9 @@ def testCoreIndex_ExtractData(nwMinimal, mockGUI):
     ]
 
     assert [(h, t) for h, t, _ in theIndex._itemIndex.iterNovelStructure(skipExcl=True)] == [
+        ("0000000000014", "T000001"),
+        ("0000000000016", "T000001"),
+        ("0000000000017", "T000001"),
         (hHandle, "T000001"),
         (sHandle, "T000001"),
         (tHandle, "T000001"),
@@ -640,6 +691,9 @@ def testCoreIndex_ExtractData(nwMinimal, mockGUI):
     # Add a fake handle to the tree and check that it's ignored
     theProject.tree._treeOrder.append("0000000000000")
     assert [(h, t) for h, t, _ in theIndex._itemIndex.iterNovelStructure(skipExcl=False)] == [
+        ("0000000000014", "T000001"),
+        ("0000000000016", "T000001"),
+        ("0000000000017", "T000001"),
         (nHandle, "T000001"),
         (nHandle, "T000011"),
         (hHandle, "T000001"),
@@ -649,25 +703,33 @@ def testCoreIndex_ExtractData(nwMinimal, mockGUI):
     theProject.tree._treeOrder.remove("0000000000000")
 
     # Extract stats
-    assert theIndex.getNovelWordCount(False) == 34
-    assert theIndex.getNovelWordCount(True) == 6
-    assert theIndex.getNovelTitleCounts(False) == [0, 2, 1, 2, 0]
-    assert theIndex.getNovelTitleCounts(True) == [0, 0, 1, 2, 0]
+    assert theIndex.getNovelWordCount(skipExcl=False) == 43
+    assert theIndex.getNovelWordCount(skipExcl=True) == 15
+    assert theIndex.getNovelTitleCounts(skipExcl=False) == [0, 3, 2, 3, 0]
+    assert theIndex.getNovelTitleCounts(skipExcl=True) == [0, 1, 2, 3, 0]
 
     # Table of Contents
-    assert theIndex.getTableOfContents(0, True) == []
-    assert theIndex.getTableOfContents(1, True) == []
-    assert theIndex.getTableOfContents(2, True) == [
+    assert theIndex.getTableOfContents(0, skipExcl=True) == []
+    assert theIndex.getTableOfContents(1, skipExcl=True) == [
+        ("0000000000014:T000001", 1, "New Novel", 15),
+    ]
+    assert theIndex.getTableOfContents(2, skipExcl=True) == [
+        ("0000000000014:T000001", 1, "New Novel", 5),
+        ("0000000000016:T000001", 2, "New Chapter", 4),
         ("%s:T000001" % hHandle, 2, "Chapter One", 6),
     ]
-    assert theIndex.getTableOfContents(3, True) == [
+    assert theIndex.getTableOfContents(3, skipExcl=True) == [
+        ("0000000000014:T000001", 1, "New Novel", 5),
+        ("0000000000016:T000001", 2, "New Chapter", 2),
+        ("0000000000017:T000001", 3, "New Scene", 2),
         ("%s:T000001" % hHandle, 2, "Chapter One", 2),
         ("%s:T000001" % sHandle, 3, "Scene One", 2),
         ("%s:T000001" % tHandle, 3, "Scene Two", 2),
     ]
 
-    assert theIndex.getTableOfContents(0, False) == []
-    assert theIndex.getTableOfContents(1, False) == [
+    assert theIndex.getTableOfContents(0, skipExcl=False) == []
+    assert theIndex.getTableOfContents(1, skipExcl=False) == [
+        ("0000000000014:T000001", 1, "New Novel", 9),
         ("%s:T000001" % nHandle, 1, "Hello World!", 12),
         ("%s:T000011" % nHandle, 1, "Hello World!", 22),
     ]
@@ -682,7 +744,9 @@ def testCoreIndex_ExtractData(nwMinimal, mockGUI):
         ("%s:T000001" % nHandle, 12), ("%s:T000011" % nHandle, 16)
     ]
 
-    assert theProject.closeProject()
+    assert theIndex.saveIndex() is True
+    assert theProject.saveProject() is True
+    assert theProject.closeProject() is True
 
     # Header Record
     bHandle = "0000000000000"
@@ -697,6 +761,156 @@ def testCoreIndex_ExtractData(nwMinimal, mockGUI):
 # END Test testCoreIndex_ExtractData
 
 
+@pytest.mark.core
+def testCoreIndex_ItemIndex(mockGUI, fncDir, mockRnd):
+    """Check the ItemIndex class.
+    """
+    theProject = NWProject(mockGUI)
+    buildTestProject(theProject, fncDir)
+
+    nHandle = "0000000000014"
+    cHandle = "0000000000016"
+    sHandle = "0000000000017"
+
+    assert theProject.index.saveIndex() is True
+    itemIndex = theProject.index._itemIndex
+
+    # The index should be empty
+    assert nHandle not in itemIndex
+    assert cHandle not in itemIndex
+    assert sHandle not in itemIndex
+
+    # Unpack Data
+    # ===========
+
+    # Data must be dictionary
+    with pytest.raises(ValueError):
+        itemIndex.unpackData("stuff")
+
+    # Keys must be valid handles
+    with pytest.raises(ValueError):
+        itemIndex.unpackData({"stuff": "more stuff"})
+
+    # Unknown keys should be skipped
+    itemIndex.unpackData({"0000000000000": {}})
+    assert itemIndex._items == {}
+
+    # Known keys can be added, even witout data
+    itemIndex.unpackData({nHandle: {}})
+    assert nHandle in itemIndex
+    itemIndex.clear()
+
+    # Add Items
+    # =========
+    assert cHandle not in itemIndex
+
+    # Add the novel chapter file
+    itemIndex.add(cHandle, theProject.tree[cHandle])
+    assert cHandle in itemIndex
+    assert itemIndex[cHandle].item == theProject.tree[cHandle]
+    assert itemIndex.mainItemHeader(cHandle) == "H0"
+    assert itemIndex.allItemTags(cHandle) == []
+    assert list(itemIndex.iterItemHeaders(cHandle))[0][0] == "T000000"
+
+    # Add a heading to the item, which should replace the T000000 heading
+    itemIndex.addItemHeading(cHandle, "T000001", "H2", "Chapter One")
+    assert itemIndex.mainItemHeader(cHandle) == "H2"
+    assert list(itemIndex.iterItemHeaders(cHandle))[0][0] == "T000001"
+
+    # Set the remainig data values
+    itemIndex.setHeadingCounts(cHandle, "T000001", 60, 10, 2)
+    itemIndex.setHeadingSynopsis(cHandle, "T000001", "In the beginning ...")
+    itemIndex.setHeadingTag(cHandle, "T000001", "One")  # Although it isn't allowed to have a tag
+    itemIndex.addHeadingReferences(cHandle, "T000001", ["Jane"], "@pov")
+    itemIndex.addHeadingReferences(cHandle, "T000001", ["Jane"], "@focus")
+    itemIndex.addHeadingReferences(cHandle, "T000001", ["Jane", "John"], "@char")
+    idxData = itemIndex.packData()
+
+    assert idxData[cHandle]["level"] == "H2"
+    assert idxData[cHandle]["headings"]["T000001"] == {
+        "level": "H2", "title": "Chapter One", "tag": "One",
+        "cCount": 60, "wCount": 10, "pCount": 2, "synopsis": "In the beginning ...",
+    }
+    assert "@pov" in idxData[cHandle]["references"]["T000001"]["Jane"]
+    assert "@focus" in idxData[cHandle]["references"]["T000001"]["Jane"]
+    assert "@char" in idxData[cHandle]["references"]["T000001"]["Jane"]
+    assert "@char" in idxData[cHandle]["references"]["T000001"]["John"]
+
+    # Add the other two files
+    itemIndex.add(nHandle, theProject.tree[nHandle])
+    itemIndex.add(sHandle, theProject.tree[sHandle])
+    itemIndex.addItemHeading(nHandle, "T000001", "H1", "Novel")
+    itemIndex.addItemHeading(sHandle, "T000001", "H3", "Scene One")
+
+    # Data Extraction
+    # ===============
+
+    # Get headers
+    allHeads = list(itemIndex.iterAllHeaders())
+    assert allHeads[0][0] == cHandle
+    assert allHeads[1][0] == nHandle
+    assert allHeads[2][0] == sHandle
+    assert allHeads[0][1] == "T000001"
+    assert allHeads[1][1] == "T000001"
+    assert allHeads[2][1] == "T000001"
+
+    # Ask for stuff that doesn't exist
+    assert itemIndex.mainItemHeader("blablabla") == "H0"
+    assert itemIndex.allItemTags("blablabla") == []
+
+    # Novel Structure
+    # ===============
+
+    # Add a second novel
+    mHandle = theProject.newRoot(nwItemClass.NOVEL)
+    uHandle = theProject.newFile("Title Page", mHandle)
+    itemIndex.add(uHandle, theProject.tree[uHandle])
+    itemIndex.addItemHeading(uHandle, "T000001", "H1", "Novel 2")
+    assert uHandle in itemIndex
+
+    # Structure of all novels
+    nStruct = list(itemIndex.iterNovelStructure())
+    assert len(nStruct) == 4
+    assert nStruct[0][0] == nHandle
+    assert nStruct[1][0] == cHandle
+    assert nStruct[2][0] == sHandle
+    assert nStruct[3][0] == uHandle
+
+    # Novel structure with root handle set
+    nStruct = list(itemIndex.iterNovelStructure(rootHandle="0000000000010"))
+    assert len(nStruct) == 3
+    assert nStruct[0][0] == nHandle
+    assert nStruct[1][0] == cHandle
+    assert nStruct[2][0] == sHandle
+
+    nStruct = list(itemIndex.iterNovelStructure(rootHandle=mHandle))
+    assert len(nStruct) == 1
+    assert nStruct[0][0] == uHandle
+
+    # Inject garbage into tree
+    theProject.tree._treeOrder.append("stuff")
+    nStruct = list(itemIndex.iterNovelStructure())
+    assert len(nStruct) == 4
+    assert nStruct[0][0] == nHandle
+    assert nStruct[1][0] == cHandle
+    assert nStruct[2][0] == sHandle
+    assert nStruct[3][0] == uHandle
+
+    # Skip excluded
+    theProject.tree[sHandle].setExported(False)
+    nStruct = list(itemIndex.iterNovelStructure(skipExcl=True))
+    assert len(nStruct) == 3
+    assert nStruct[0][0] == nHandle
+    assert nStruct[1][0] == cHandle
+    assert nStruct[2][0] == uHandle
+
+    # Delete new item
+    del itemIndex[uHandle]
+    assert uHandle not in itemIndex
+
+# END Test testCoreIndex_ItemIndex
+
+
 @pytest.mark.core
 def testCoreIndex_CountWords():
     """Test the word counter and the exclusion filers.