diff --git a/sqlite_utils/db.py b/sqlite_utils/db.py
index f43d937..4ff64c5 100644
--- a/sqlite_utils/db.py
+++ b/sqlite_utils/db.py
@@ -901,15 +901,15 @@ class Table(Queryable):
             for col, typ in self.columns_dict.items()
             if col in columns
         }
-        if lookup_table.exists() and not set(lookup_columns_definition.keys()).issubset(
-            lookup_table.columns_dict.keys()
-        ):
-            # TODO: Write test for this
-            raise InvalidColumns(
-                "Lookup table {} already exists but does not have columns {}".format(
-                    table, lookup_columns_definition.keys()
+        if lookup_table.exists():
+            if not set(lookup_columns_definition.items()).issubset(
+                lookup_table.columns_dict.items()
+            ):
+                raise InvalidColumns(
+                    "Lookup table {} already exists but does not have columns {}".format(
+                        table, lookup_columns_definition
+                    )
                 )
-            )
         else:
             lookup_table.create(
                 {
@@ -921,9 +921,9 @@ class Table(Queryable):
                 pk="id",
             )
         lookup_columns = [(rename.get(col) or col) for col in columns]
-        lookup_table.create_index(lookup_columns, unique=True)
+        lookup_table.create_index(lookup_columns, unique=True, if_not_exists=True)
         self.db.execute(
-            "INSERT INTO [{lookup_table}] ({lookup_columns}) SELECT DISTINCT {table_cols} FROM [{table}]".format(
+            "INSERT OR IGNORE INTO [{lookup_table}] ({lookup_columns}) SELECT DISTINCT {table_cols} FROM [{table}]".format(
                 lookup_table=table,
                 lookup_columns=", ".join("[{}]".format(c) for c in lookup_columns),
                 table_cols=", ".join("[{}]".format(c) for c in columns),
diff --git a/tests/test_extract.py b/tests/test_extract.py
index c9e94d3..25dd6e2 100644
--- a/tests/test_extract.py
+++ b/tests/test_extract.py
@@ -130,3 +130,44 @@ def test_extract_rowid_table(fresh_db):
         "   FOREIGN KEY(common_name_latin_name_id) REFERENCES common_name_latin_name(id)\n"
         ")"
     )
+
+
+def test_reuse_lookup_table(fresh_db):
+    fresh_db["species"].insert({"id": 1, "name": "Wolf"}, pk="id")
+    fresh_db["sightings"].insert({"id": 10, "species": "Wolf"}, pk="id")
+    fresh_db["individuals"].insert(
+        {"id": 10, "name": "Terriana", "species": "Fox"}, pk="id"
+    )
+    fresh_db["sightings"].extract("species", rename={"species": "name"})
+    fresh_db["individuals"].extract("species", rename={"species": "name"})
+    assert fresh_db["sightings"].schema == (
+        'CREATE TABLE "sightings" (\n'
+        "   [id] INTEGER PRIMARY KEY,\n"
+        "   [species_id] INTEGER,\n"
+        "   FOREIGN KEY(species_id) REFERENCES species(id)\n"
+        ")"
+    )
+    assert fresh_db["individuals"].schema == (
+        'CREATE TABLE "individuals" (\n'
+        "   [id] INTEGER PRIMARY KEY,\n"
+        "   [name] TEXT,\n"
+        "   [species_id] INTEGER,\n"
+        "   FOREIGN KEY(species_id) REFERENCES species(id)\n"
+        ")"
+    )
+    assert list(fresh_db["species"].rows) == [
+        {"id": 1, "name": "Wolf"},
+        {"id": 2, "name": "Fox"},
+    ]
+
+
+def test_extract_error_on_incompatible_existing_lookup_table(fresh_db):
+    fresh_db["species"].insert({"id": 1})
+    fresh_db["tree"].insert({"name": "Tree 1", "common_name": "Palm"})
+    with pytest.raises(InvalidColumns):
+        fresh_db["tree"].extract("common_name", table="species")
+
+    # Try again with incompatible existing column type
+    fresh_db["species2"].insert({"id": 1, "common_name": 3.5})
+    with pytest.raises(InvalidColumns):
+        fresh_db["tree"].extract("common_name", table="species2")