diff --git a/CHANGELOG.md b/CHANGELOG.md
index 961fe01a3..49e030aba 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -3,6 +3,9 @@
 ### Fixes
 - `adapter.get_columns_in_relation` (method) and `get_columns_in_relation` (macro) now return identical responses. The previous behavior of `get_columns_in_relation` (macro) is now represented by a new macro, `get_columns_in_relation_raw` ([#354](https://github.com/dbt-labs/dbt-spark/issues/354), [#355](https://github.com/dbt-labs/dbt-spark/pull/355))
 
+### Under the hood
+- Add `DBT_INVOCATION_ENV` environment variable to ODBC user agent string ([#366](https://github.com/dbt-labs/dbt-spark/pull/366))
+
 ## dbt-spark 1.1.0 (April 28, 2022)
 
 ### Features
diff --git a/dbt/adapters/spark/connections.py b/dbt/adapters/spark/connections.py
index 59ceb9dd8..135463eb3 100644
--- a/dbt/adapters/spark/connections.py
+++ b/dbt/adapters/spark/connections.py
@@ -1,3 +1,5 @@
+import os
+
 from contextlib import contextmanager
 
 import dbt.exceptions
@@ -7,6 +9,7 @@
 from dbt.events import AdapterLogger
 from dbt.utils import DECIMALS
 from dbt.adapters.spark import __version__
+from dbt.tracking import DBT_INVOCATION_ENV
 
 try:
     from TCLIService.ttypes import TOperationState as ThriftState
@@ -409,9 +412,8 @@ def open(cls, connection):
                     cls.validate_creds(creds, required_fields)
 
                     dbt_spark_version = __version__.version
-                    user_agent_entry = (
-                        f"dbt-labs-dbt-spark/{dbt_spark_version} (Databricks)"  # noqa
-                    )
+                    dbt_invocation_env = os.getenv(DBT_INVOCATION_ENV) or "manual"
+                    user_agent_entry = f"dbt-labs-dbt-spark/{dbt_spark_version} (Databricks, {dbt_invocation_env})"  # noqa
 
                     # http://simba.wpengine.com/products/Spark/doc/ODBC_InstallGuide/unix/content/odbc/hi/configuring/serverside.htm
                     ssp = {f"SSP_{k}": f"{{{v}}}" for k, v in creds.server_side_parameters.items()}