diff --git a/docs/en_US/release_notes_3_7.rst b/docs/en_US/release_notes_3_7.rst
index e8af0f963..6af835ed6 100644
--- a/docs/en_US/release_notes_3_7.rst
+++ b/docs/en_US/release_notes_3_7.rst
@@ -32,5 +32,6 @@ Bug fixes
 | `Bug #3798 <https://redmine.postgresql.org/issues/3798>`_ - Ensure the browser toolbar buttons work in languages other than English.
 | `Bug #3805 <https://redmine.postgresql.org/issues/3805>`_ - Allow horizontal sizing of the edit grid text pop-out.
 | `Bug #3809 <https://redmine.postgresql.org/issues/3809>`_ - Ensure auto complete should works when first identifier in the FROM clause needs quoting.
+| `Bug #3810 <https://redmine.postgresql.org/issues/3810>`_ - Ensure auto complete should works for columns from a schema-qualified table.
 | `Bug #3821 <https://redmine.postgresql.org/issues/3821>`_ - Ensure identifiers are properly displayed in the plan viewer.
 | `Bug #3830 <https://redmine.postgresql.org/issues/3830>`_ - Make the setup process more robust against aborted executions.
\ No newline at end of file
diff --git a/web/pgadmin/utils/sqlautocomplete/autocomplete.py b/web/pgadmin/utils/sqlautocomplete/autocomplete.py
index 502deb7cf..4ad928c7d 100644
--- a/web/pgadmin/utils/sqlautocomplete/autocomplete.py
+++ b/web/pgadmin/utils/sqlautocomplete/autocomplete.py
@@ -440,7 +440,7 @@ class SQLAutoComplete(object):
                 # Text starts with double quote; Remove quoting and
                 # match on everything that follows the double-quote.
                 item = self.unescape_name(item.lower())
-                match_point = item.lower().find(text, 0, match_end_limit)
+                match_point = item.find(text, 0, match_end_limit)
                 if match_point >= 0:
                     # Use negative infinity to force keywords to sort after all
                     # fuzzy matches
@@ -525,10 +525,15 @@ class SQLAutoComplete(object):
         return result
 
     def get_column_matches(self, suggestion, word_before_cursor):
+        schema = None
+        if len(suggestion.table_refs) > 0 and \
+                hasattr(suggestion.table_refs[0], 'schema') and \
+                suggestion.table_refs[0].schema != '':
+            schema = suggestion.table_refs[0].schema
 
         # Tables and Views should be populated first.
-        self.fetch_schema_objects(None, 'tables')
-        self.fetch_schema_objects(None, 'views')
+        self.fetch_schema_objects(schema, 'tables')
+        self.fetch_schema_objects(schema, 'views')
 
         tables = suggestion.table_refs
         do_qualify = suggestion.qualifiable and {