diff --git a/docs/table.rst b/docs/table.rst
index 80e006a..df060ed 100644
--- a/docs/table.rst
+++ b/docs/table.rst
@@ -204,3 +204,21 @@ If you insert additional records into the table you will need to refresh the sea
         "is_good_dog": True,
     }, pk="id")
     dogs.populate_fts(["name", "twitter"])
+
+Creating indexes
+================
+
+You can create an index on a table using the ``.create_index(columns)`` method. The method takes a list of columns:
+
+.. code-block:: python
+
+    dogs.create_index(["is_good_dog"])
+
+By default the index will be named ``idx_{table-name}_{columns}`` - if you want to customize the name of the created index you can pass the ``index_name`` parameter::]
+
+.. code-block:: python
+
+    dogs.create_index(
+        ["is_good_dog", "age"],
+        index_name="good_dogs_by_age"
+    )
diff --git a/sqlite_utils/db.py b/sqlite_utils/db.py
index 3f922c0..82f1f93 100644
--- a/sqlite_utils/db.py
+++ b/sqlite_utils/db.py
@@ -140,6 +140,19 @@ class Table:
         self.db.create_table(self.name, columns, pk=pk, foreign_keys=foreign_keys)
         self.exists = True
 
+    def create_index(self, columns, index_name=None):
+        if index_name is None:
+            index_name = "idx_{}_{}".format(
+                self.name.replace(" ", "_"), "_".join(columns)
+            )
+        sql = """
+            CREATE INDEX {index_name}
+                ON {table_name} ({columns});
+        """.format(
+            index_name=index_name, table_name=self.name, columns=", ".join(columns)
+        )
+        self.db.conn.execute(sql)
+
     def drop(self):
         return self.db.conn.execute("DROP TABLE {}".format(self.name))
 
diff --git a/tests/test_create.py b/tests/test_create.py
index 3ae5fef..966602b 100644
--- a/tests/test_create.py
+++ b/tests/test_create.py
@@ -1,4 +1,5 @@
 from .fixtures import fresh_db
+from sqlite_utils.db import Index
 import pytest
 import json
 
@@ -68,6 +69,55 @@ def test_create_table_works_for_m2m_with_only_foreign_keys(fresh_db):
     )
 
 
+@pytest.mark.parametrize(
+    "columns,index_name,expected_index",
+    (
+        (
+            ["is_good_dog"],
+            None,
+            Index(
+                seq=0,
+                name="idx_dogs_is_good_dog",
+                unique=0,
+                origin="c",
+                partial=0,
+                columns=["is_good_dog"],
+            ),
+        ),
+        (
+            ["is_good_dog", "age"],
+            None,
+            Index(
+                seq=0,
+                name="idx_dogs_is_good_dog_age",
+                unique=0,
+                origin="c",
+                partial=0,
+                columns=["is_good_dog", "age"],
+            ),
+        ),
+        (
+            ["age"],
+            "age_index",
+            Index(
+                seq=0,
+                name="age_index",
+                unique=0,
+                origin="c",
+                partial=0,
+                columns=["age"],
+            ),
+        ),
+    ),
+)
+def test_create_index(fresh_db, columns, index_name, expected_index):
+    dogs = fresh_db["dogs"]
+    dogs.insert({"name": "Cleo", "twitter": "cleopaws", "age": 3, "is_good_dog": True})
+    assert [] == dogs.indexes
+    dogs.create_index(columns, index_name)
+    assert expected_index == dogs.indexes[0]
+
+
 @pytest.mark.parametrize(
     "data_structure",
     (