diff --git a/sqlglot/generators/exasol.py b/sqlglot/generators/exasol.py
index a729471af5..d560603dfb 100644
--- a/sqlglot/generators/exasol.py
+++ b/sqlglot/generators/exasol.py
@@ -14,6 +14,7 @@
 )
 from sqlglot.errors import UnsupportedError
 from sqlglot.generator import unsupported_args
+from sqlglot.optimizer.annotate_types import annotate_types
 from sqlglot.optimizer.scope import build_scope
 from sqlglot.parsers.exasol import DATE_UNITS
 
@@ -398,6 +399,7 @@ def datatype_sql(self, expression: exp.DataType) -> str:
             ]
         ),
         exp.SubstringIndex: _substring_index_sql,
+        exp.JSONObject: lambda self, e: self.jsonobject_sql(e),
         exp.WeekOfYear: rename_func("WEEK"),
         # https://docs.exasol.com/db/latest/sql_references/functions/alphabeticallistfunctions/to_date.htm
         exp.Date: rename_func("TO_DATE"),
@@ -929,6 +931,61 @@ def jsonextract_sql(self, expression: exp.JSONExtract) -> str:
 
         return sql
 
+    def jsonobject_sql(self, expression: exp.JSONObject) -> str:
+        pairs = expression.expressions or []
+        if not pairs:
+            return self.sql(exp.Literal.string("{}"))
+
+        concat_args: list[exp.Expression] = [exp.Literal.string("{")]
+        for i, pair in enumerate(pairs):
+            if not isinstance(pair, exp.JSONKeyValue):
+                return self.function_fallback_sql(expression)
+            key = pair.this
+            value = pair.args.get("expression")
+            if key is None or value is None:
+                return self.function_fallback_sql(expression)
+
+            prefix = ", " if i > 0 else ""
+            key_label = f'{prefix}"{key.name.replace(chr(34), chr(92) + chr(34))}": '
+            concat_args.append(exp.Literal.string(key_label))
+
+            if value.is_string:
+                wrapped: exp.Expression = exp.Concat(
+                    expressions=[
+                        exp.Literal.string('"'),
+                        value,
+                        exp.Literal.string('"'),
+                    ]
+                )
+            else:
+                typed_value = value if value.type else annotate_types(value, dialect=self.dialect)
+                if typed_value.is_type(*exp.DataType.TEXT_TYPES):
+                    wrapped = exp.Case(
+                        ifs=[
+                            exp.If(
+                                this=typed_value.is_(exp.Null()),
+                                true=exp.Literal.string("null"),
+                            )
+                        ],
+                        default=exp.Concat(
+                            expressions=[
+                                exp.Literal.string('"'),
+                                typed_value.copy(),
+                                exp.Literal.string('"'),
+                            ]
+                        ),
+                    )
+                else:
+                    wrapped = exp.Coalesce(
+                        this=exp.cast(typed_value, exp.DataType.build("VARCHAR(100)")),
+                        expressions=[exp.Literal.string("null")],
+                    )
+
+            concat_args.append(wrapped)
+
+        concat_args.append(exp.Literal.string("}"))
+        return self.sql(exp.Concat(expressions=concat_args))
+
     @unsupported_args("flag")
     def regexplike_sql(self, expression: exp.RegexpLike) -> str:
         if not expression.args.get("full_match"):
diff --git a/tests/dialects/test_exasol.py b/tests/dialects/test_exasol.py
index 3dabcf07c8..5daeeb80e4 100644
--- a/tests/dialects/test_exasol.py
+++ b/tests/dialects/test_exasol.py
@@ -991,3 +991,52 @@ def test_show_tables(self):
                 "exasol": "SELECT TABLE_NAME FROM SYS.EXA_ALL_TABLES WHERE TABLE_SCHEMA = CURRENT_SCHEMA"
             },
         )
+
+    def test_json_object(self):
+        from sqlglot import parse_one
+        from sqlglot.optimizer.annotate_types import annotate_types
+        from sqlglot.optimizer.qualify import qualify
+
+        # Empty JSON_OBJECT -> '{}' literal
+        self.validate_all(
+            "SELECT '{}'",
+            read={"mysql": "SELECT JSON_OBJECT()"},
+            write={"exasol": "SELECT '{}'"},
+        )
+
+        # String-typed column: NULL-safe quoted value via CASE
+        ast = parse_one("SELECT JSON_OBJECT('name', str_name) AS j FROM t", read="mysql")
+        schema = {"t": {"str_name": "VARCHAR"}}
+        annotated = annotate_types(
+            qualify(ast, schema=schema, dialect="mysql"),
+            schema=schema,
+            dialect="mysql",
+        )
+        result = annotated.sql("exasol")
+        self.assertIn("CASE WHEN", result)
+        self.assertIn("IS NULL THEN 'null'", result)
+        self.assertIn("'\"name\": '", result)
+
+        # Numeric-typed column: COALESCE(CAST(..) AS VARCHAR(100)), 'null')
+        ast = parse_one("SELECT JSON_OBJECT('id', int_id) AS j FROM t", read="mysql")
+        schema = {"t": {"int_id": "INT"}}
+        annotated = annotate_types(
+            qualify(ast, schema=schema, dialect="mysql"),
+            schema=schema,
+            dialect="mysql",
+        )
+        result = annotated.sql("exasol")
+        self.assertIn("COALESCE(CAST(", result)
+        self.assertIn("AS VARCHAR(100)), 'null')", result)
+
+        # Multi-pair with comma separators
+        ast = parse_one("SELECT JSON_OBJECT('id', int_id, 'name', str_name) FROM t", read="mysql")
+        schema = {"t": {"int_id": "INT", "str_name": "VARCHAR"}}
+        annotated = annotate_types(
+            qualify(ast, schema=schema, dialect="mysql"),
+            schema=schema,
+            dialect="mysql",
+        )
+        result = annotated.sql("exasol")
+        self.assertIn("'\"id\": '", result)
+        self.assertIn("', \"name\": '", result)