diff --git a/docs/help.md b/docs/help.md
index eafe8ad..9f089bb 100644
--- a/docs/help.md
+++ b/docs/help.md
@@ -147,6 +147,8 @@ Options:
 
 Commands:
   list*   Show recent logged prompts and their responses
+  off     Turn off logging for all prompts
+  on      Turn on logging for all prompts
   path    Output the path to the logs.db file
   status  Show current status of database logging
 ```
@@ -165,6 +167,24 @@ Usage: llm logs status [OPTIONS]
 
   Show current status of database logging
 
+Options:
+  --help  Show this message and exit.
+```
+#### llm logs on --help
+```
+Usage: llm logs on [OPTIONS]
+
+  Turn on logging for all prompts
+
+Options:
+  --help  Show this message and exit.
+```
+#### llm logs off --help
+```
+Usage: llm logs off [OPTIONS]
+
+  Turn off logging for all prompts
+
 Options:
   --help  Show this message and exit.
 ```
diff --git a/llm/cli.py b/llm/cli.py
index f0500ae..d30e025 100644
--- a/llm/cli.py
+++ b/llm/cli.py
@@ -247,15 +247,11 @@ def prompt(
         raise click.ClickException(str(ex))
 
     # Log to the database
-    if no_log:
-        return
-
-    log_path = logs_db_path()
-    if not log_path.exists():
-        return
-    db = sqlite_utils.Database(log_path)
-    migrate(db)
-    response.log_to_db(db)
+    if logs_on():
+        log_path = logs_db_path()
+        db = sqlite_utils.Database(log_path)
+        migrate(db)
+        response.log_to_db(db)
 
 
 def load_conversation(conversation_id: Optional[str]) -> Optional[Conversation]:
@@ -358,6 +354,10 @@ def logs_status():
     if not path.exists():
         click.echo("No log database found at {}".format(path))
         return
+    if logs_on():
+        click.echo("Logging is ON for all prompts".format())
+    else:
+        click.echo("Logging is OFF".format())
     db = sqlite_utils.Database(path)
     migrate(db)
     click.echo("Found log database at {}".format(path))
@@ -368,6 +368,21 @@ def logs_status():
     )
 
 
+@logs.command(name="on")
+def logs_turn_on():
+    "Turn on logging for all prompts"
+    path = user_dir() / "logs-off"
+    if path.exists():
+        path.unlink()
+
+
+@logs.command(name="off")
+def logs_turn_off():
+    "Turn off logging for all prompts"
+    path = user_dir() / "logs-off"
+    path.touch()
+
+
 @logs.command(name="list")
 @click.option(
     "-n",
@@ -687,3 +702,7 @@ def _human_readable_size(size_bytes):
         i += 1
 
     return "{:.2f}{}".format(size_bytes, size_name[i])
+
+
+def logs_on():
+    return not (user_dir() / "logs-off").exists()
diff --git a/tests/test_llm.py b/tests/test_llm.py
index 0d7a381..e93b583 100644
--- a/tests/test_llm.py
+++ b/tests/test_llm.py
@@ -65,11 +65,25 @@ def test_logs_path(monkeypatch, env, user_path):
 
 @mock.patch.dict(os.environ, {"OPENAI_API_KEY": "X"})
 @pytest.mark.parametrize("use_stdin", (True, False))
-def test_llm_default_prompt(mocked_openai, use_stdin, user_path):
+@pytest.mark.parametrize("logs_off", (True, False))
+def test_llm_default_prompt(mocked_openai, use_stdin, user_path, logs_off):
     # Reset the log_path database
     log_path = user_path / "logs.db"
     log_db = sqlite_utils.Database(str(log_path))
     log_db["responses"].delete_where()
+
+    logs_off_path = user_path / "logs-off"
+    if logs_off:
+        # Turn off logging
+        assert not logs_off_path.exists()
+        CliRunner().invoke(cli, ["logs", "off"])
+        assert logs_off_path.exists()
+    else:
+        # Turn on logging
+        CliRunner().invoke(cli, ["logs", "on"])
+        assert not logs_off_path.exists()
+
+    # Run the prompt
     runner = CliRunner()
     prompt = "three names for a pet pelican"
     input = None
@@ -85,6 +99,11 @@ def test_llm_default_prompt(mocked_openai, use_stdin, user_path):
 
     # Was it logged?
     rows = list(log_db["responses"].rows)
+
+    if logs_off:
+        assert len(rows) == 0
+        return
+
     assert len(rows) == 1
     expected = {
         "model": "gpt-3.5-turbo",
@@ -109,6 +128,8 @@ def test_llm_default_prompt(mocked_openai, use_stdin, user_path):
     # Test "llm logs"
     log_result = runner.invoke(cli, ["logs", "-n", "1"], catch_exceptions=False)
     log_json = json.loads(log_result.output)
+
+    # Should have logged correctly:
     assert (
         log_json[0].items()
         >= {