diff --git a/.changes/unreleased/Under the Hood-20240612-195629.yaml b/.changes/unreleased/Under the Hood-20240612-195629.yaml
new file mode 100644
index 000000000..c90ebcdab
--- /dev/null
+++ b/.changes/unreleased/Under the Hood-20240612-195629.yaml	
@@ -0,0 +1,6 @@
+kind: Under the Hood
+body: Lazy load agate to improve performance
+time: 2024-06-12T19:56:29.943204-07:00
+custom:
+  Author: versusfacit
+  Issue: "1049"
diff --git a/dbt/adapters/spark/impl.py b/dbt/adapters/spark/impl.py
index 255ab7806..d33ebde20 100644
--- a/dbt/adapters/spark/impl.py
+++ b/dbt/adapters/spark/impl.py
@@ -14,6 +14,7 @@
     Callable,
     Set,
     FrozenSet,
+    TYPE_CHECKING,
 )
 
 from dbt.adapters.base.relation import InformationSchema
@@ -24,7 +25,10 @@
 
 from typing_extensions import TypeAlias
 
-import agate
+if TYPE_CHECKING:
+    # Indirectly imported via agate_helper, which is lazy loaded further downfile.
+    # Used by mypy for earlier type hints.
+    import agate
 
 from dbt.adapters.base import AdapterConfig, PythonJobHelper
 from dbt.adapters.base.impl import catch_as_completed, ConstraintSupport
@@ -127,34 +131,36 @@ def date_function(cls) -> str:
         return "current_timestamp()"
 
     @classmethod
-    def convert_text_type(cls, agate_table: agate.Table, col_idx: int) -> str:
+    def convert_text_type(cls, agate_table: "agate.Table", col_idx: int) -> str:
         return "string"
 
     @classmethod
-    def convert_number_type(cls, agate_table: agate.Table, col_idx: int) -> str:
+    def convert_number_type(cls, agate_table: "agate.Table", col_idx: int) -> str:
+        import agate
+
         decimals = agate_table.aggregate(agate.MaxPrecision(col_idx))
         return "double" if decimals else "bigint"
 
     @classmethod
-    def convert_integer_type(cls, agate_table: agate.Table, col_idx: int) -> str:
+    def convert_integer_type(cls, agate_table: "agate.Table", col_idx: int) -> str:
         return "bigint"
 
     @classmethod
-    def convert_date_type(cls, agate_table: agate.Table, col_idx: int) -> str:
+    def convert_date_type(cls, agate_table: "agate.Table", col_idx: int) -> str:
         return "date"
 
     @classmethod
-    def convert_time_type(cls, agate_table: agate.Table, col_idx: int) -> str:
+    def convert_time_type(cls, agate_table: "agate.Table", col_idx: int) -> str:
         return "time"
 
     @classmethod
-    def convert_datetime_type(cls, agate_table: agate.Table, col_idx: int) -> str:
+    def convert_datetime_type(cls, agate_table: "agate.Table", col_idx: int) -> str:
         return "timestamp"
 
     def quote(self, identifier: str) -> str:
         return "`{}`".format(identifier)
 
-    def _get_relation_information(self, row: agate.Row) -> RelationInfo:
+    def _get_relation_information(self, row: "agate.Row") -> RelationInfo:
         """relation info was fetched with SHOW TABLES EXTENDED"""
         try:
             _schema, name, _, information = row
@@ -165,7 +171,7 @@ def _get_relation_information(self, row: agate.Row) -> RelationInfo:
 
         return _schema, name, information
 
-    def _get_relation_information_using_describe(self, row: agate.Row) -> RelationInfo:
+    def _get_relation_information_using_describe(self, row: "agate.Row") -> RelationInfo:
         """Relation info fetched using SHOW TABLES and an auxiliary DESCRIBE statement"""
         try:
             _schema, name, _ = row
@@ -193,8 +199,8 @@ def _get_relation_information_using_describe(self, row: agate.Row) -> RelationIn
 
     def _build_spark_relation_list(
         self,
-        row_list: agate.Table,
-        relation_info_func: Callable[[agate.Row], RelationInfo],
+        row_list: "agate.Table",
+        relation_info_func: Callable[["agate.Row"], RelationInfo],
     ) -> List[BaseRelation]:
         """Aggregate relations with format metadata included."""
         relations = []
@@ -370,7 +376,7 @@ def get_catalog(
         self,
         relation_configs: Iterable[RelationConfig],
         used_schemas: FrozenSet[Tuple[str, str]],
-    ) -> Tuple[agate.Table, List[Exception]]:
+    ) -> Tuple["agate.Table", List[Exception]]:
         schema_map = self._get_catalog_schemas(relation_configs)
         if len(schema_map) > 1:
             raise CompilationError(
@@ -378,7 +384,7 @@ def get_catalog(
             )
 
         with executor(self.config) as tpe:
-            futures: List[Future[agate.Table]] = []
+            futures: List[Future["agate.Table"]] = []
             for info, schemas in schema_map.items():
                 for schema in schemas:
                     futures.append(
@@ -399,7 +405,7 @@ def _get_one_catalog(
         information_schema: InformationSchema,
         schemas: Set[str],
         used_schemas: FrozenSet[Tuple[str, str]],
-    ) -> agate.Table:
+    ) -> "agate.Table":
         if len(schemas) != 1:
             raise CompilationError(
                 f"Expected only one schema in spark _get_one_catalog, found " f"{schemas}"
@@ -412,6 +418,9 @@ def _get_one_catalog(
         for relation in self.list_relations(database, schema):
             logger.debug("Getting table schema for relation {}", str(relation))
             columns.extend(self._get_columns_for_catalog(relation))
+
+        import agate
+
         return agate.Table.from_object(columns, column_types=DEFAULT_TYPE_TESTER)
 
     def check_schema_exists(self, database: str, schema: str) -> bool:
@@ -486,7 +495,7 @@ def python_submission_helpers(self) -> Dict[str, Type[PythonJobHelper]]:
             "all_purpose_cluster": AllPurposeClusterPythonJobHelper,
         }
 
-    def standardize_grants_dict(self, grants_table: agate.Table) -> dict:
+    def standardize_grants_dict(self, grants_table: "agate.Table") -> dict:
         grants_dict: Dict[str, List[str]] = {}
         for row in grants_table:
             grantee = row["Principal"]