diff --git a/isaaclab_arena/relations/object_placer.py b/isaaclab_arena/relations/object_placer.py
index d8d279b18..2515a6e30 100644
--- a/isaaclab_arena/relations/object_placer.py
+++ b/isaaclab_arena/relations/object_placer.py
@@ -11,7 +11,7 @@
 
 from isaaclab_arena.relations.bounding_box_helpers import assign_variants_for_envs, build_per_env_bounding_boxes
 from isaaclab_arena.relations.object_placer_params import ObjectPlacerParams
-from isaaclab_arena.relations.placement_result import MultiEnvPlacementResult, PlacementResult
+from isaaclab_arena.relations.placement_result import PlacementResult
 from isaaclab_arena.relations.placement_validation import PlacementCheck, PlacementValidationResults
 from isaaclab_arena.relations.relation_solver import RelationSolver
 from isaaclab_arena.relations.relations import (
@@ -78,7 +78,7 @@ def place(
         self,
         objects: list[ObjectBase],
         num_envs: int = 1,
-    ) -> PlacementResult | MultiEnvPlacementResult:
+    ) -> list[PlacementResult]:
         """Place objects according to their spatial relations.
 
         Every environment is solved against its own per-env bounding boxes and
@@ -93,8 +93,7 @@ def place(
                 placement (one layout per env).
 
         Returns:
-            PlacementResult when num_envs == 1, otherwise a
-            MultiEnvPlacementResult with one layout per environment.
+            One PlacementResult per environment.
         """
         anchor_objects_set, generator = self._prepare_placement(objects)
         max_attempts = self.params.max_placement_attempts
@@ -123,9 +122,7 @@ def place(
             orientations_per_env = [r.orientations for r in results_per_env]
             self._apply_poses(positions_per_env, anchor_objects_set, orientations_per_env)
 
-        if num_envs == 1:
-            return results_per_env[0]
-        return MultiEnvPlacementResult(results=results_per_env)
+        return results_per_env
 
     def place_ranked_per_env(
         self,
@@ -654,7 +651,7 @@ def _validate_placement(
             env_bboxes: Per-object bboxes for the current env, each with shape (1, 3).
 
         Returns:
-            PlacementValidationResults containing the validation results.
+            PlacementValidationResults with the overlap and on-relation checks.
         """
         no_overlap = self._validate_no_overlap(positions, env_bboxes)
         on_relation = self._validate_on_relations(positions, env_bboxes)
diff --git a/isaaclab_arena/relations/placement_result.py b/isaaclab_arena/relations/placement_result.py
index 6f79b9e80..13cd9b252 100644
--- a/isaaclab_arena/relations/placement_result.py
+++ b/isaaclab_arena/relations/placement_result.py
@@ -15,7 +15,7 @@
 
 @dataclass
 class PlacementResult:
-    """Result of an ObjectPlacer.place() call."""
+    """Solved object layout for one environment."""
 
     validation_results: PlacementValidationResults
     """Validation checklist for the placement."""
@@ -35,28 +35,8 @@ class PlacementResult:
 
     @property
     def success(self) -> bool:
-        """True when this layout passed every validation check.
+        """True when all required validation checks pass.
 
-        Soft selection: place() always returns the best-ranked layout per env, even when no
-        candidate validated. Callers check success to distinguish a validated layout from a
-        lowest-loss fallback; failed_items on the checklist says which checks failed.
+        place() returns a best-loss fallback even on failure; check this to tell validated from fallback.
         """
         return self.validation_results.do_all_required_validation_checks_pass()
-
-
-@dataclass
-class MultiEnvPlacementResult:
-    """Result of an ObjectPlacer.place() call for multiple environments."""
-
-    results: list[PlacementResult]
-    """One PlacementResult per environment (same length as num_envs)."""
-
-    @property
-    def success(self) -> bool:
-        """True if every environment's placement succeeded."""
-        return all(r.success for r in self.results)
-
-    @property
-    def attempts(self) -> int:
-        """Number of attempts (same for all envs in the batched run)."""
-        return self.results[0].attempts if self.results else 0
diff --git a/isaaclab_arena/relations/pooled_object_placer.py b/isaaclab_arena/relations/pooled_object_placer.py
index d13211b55..640195743 100644
--- a/isaaclab_arena/relations/pooled_object_placer.py
+++ b/isaaclab_arena/relations/pooled_object_placer.py
@@ -12,7 +12,7 @@
 from isaaclab_arena.relations.bounding_box_helpers import has_heterogeneous_objects
 from isaaclab_arena.relations.object_placer import ObjectPlacer
 from isaaclab_arena.relations.object_placer_params import ObjectPlacerParams
-from isaaclab_arena.relations.placement_result import MultiEnvPlacementResult, PlacementResult
+from isaaclab_arena.relations.placement_result import PlacementResult
 from isaaclab_arena.utils.random import get_rngs
 
 if TYPE_CHECKING:
@@ -186,10 +186,7 @@ def _solve_reusable_layouts(self, num_layouts: int, allow_fallback: bool = False
         """
         self._prepare_seeded_solve(num_layouts * self._placer.params.max_placement_attempts)
         with torch.inference_mode(False):
-            result = self._placer.place(self._objects, num_envs=num_layouts)
-
-        # place() returns a single PlacementResult only when num_envs == 1.
-        all_layouts = result.results if isinstance(result, MultiEnvPlacementResult) else [result]
+            all_layouts = self._placer.place(self._objects, num_envs=num_layouts)
         valid_layouts = [layout for layout in all_layouts if layout.success]
 
         if len(valid_layouts) < num_layouts:
diff --git a/isaaclab_arena/tests/test_heterogeneous_placement.py b/isaaclab_arena/tests/test_heterogeneous_placement.py
index 5392b3e9a..49396c40c 100644
--- a/isaaclab_arena/tests/test_heterogeneous_placement.py
+++ b/isaaclab_arena/tests/test_heterogeneous_placement.py
@@ -17,7 +17,7 @@
 from isaaclab_arena.relations.bounding_box_helpers import build_per_env_bounding_boxes, get_bounding_box_per_env
 from isaaclab_arena.relations.object_placer import ObjectPlacer
 from isaaclab_arena.relations.object_placer_params import ObjectPlacerParams
-from isaaclab_arena.relations.placement_result import MultiEnvPlacementResult, PlacementResult
+from isaaclab_arena.relations.placement_result import PlacementResult
 from isaaclab_arena.relations.placement_validation import PlacementValidationResults
 from isaaclab_arena.relations.pooled_object_placer import PooledObjectPlacer
 from isaaclab_arena.relations.relation_solver import RelationSolver
@@ -252,9 +252,8 @@ def test_object_placer_heterogeneous_produces_per_env_results():
     placer = ObjectPlacer(params=params)
     result = placer.place(objects, num_envs=num_envs)
 
-    assert isinstance(result, MultiEnvPlacementResult)
-    assert len(result.results) == num_envs
-    for r in result.results:
+    assert len(result) == num_envs
+    for r in result:
         assert hetero_box in r.positions
 
 
@@ -283,10 +282,8 @@ def test_object_placer_heterogeneous_z_height_matches_variant():
     placer = ObjectPlacer(params=params)
     result = placer.place(objects, num_envs=num_envs)
 
-    assert isinstance(result, MultiEnvPlacementResult)
-    # Both envs should have solved z near the desk top + clearance (0.11).
-    # The On loss targets: z = parent_top + clearance - child_min_z = 0.1 + 0.01 - 0.0 = 0.11
-    for env_idx, r in enumerate(result.results):
+    assert len(result) == num_envs
+    for env_idx, r in enumerate(result):
         z = r.positions[hetero][2]
         assert abs(z - 0.11) < 0.05, f"Env {env_idx}: z={z:.4f}, expected ~0.11"
 
@@ -329,10 +326,9 @@ def test_mixed_heterogeneous_and_homogeneous_placement():
     placer = ObjectPlacer(params=params)
     result = placer.place(objects, num_envs=num_envs)
 
-    assert isinstance(result, MultiEnvPlacementResult)
-    assert len(result.results) == num_envs
+    assert len(result) == num_envs
 
-    for env_idx, r in enumerate(result.results):
+    for env_idx, r in enumerate(result):
         assert obj_a in r.positions and obj_x in r.positions
         # Verify z-height is near desk top + clearance for both objects.
         for obj in (obj_a, obj_x):
@@ -365,8 +361,7 @@ def test_heterogeneous_placement_always_returns_per_env_results():
     placer = ObjectPlacer(params=params)
     result = placer.place([desk, hetero], num_envs=4)
 
-    assert isinstance(result, MultiEnvPlacementResult)
-    assert len(result.results) == 4
+    assert len(result) == 4
 
 
 def test_object_placer_place_ranked_per_env_returns_sorted_env_lists():
@@ -424,9 +419,8 @@ def test_object_placer_homogeneous_objects_return_multi_env_result():
     placer = ObjectPlacer(params=params)
     result = placer.place([desk, box], num_envs=2)
 
-    assert isinstance(result, MultiEnvPlacementResult)
-    assert len(result.results) == 2
-    assert all(r.success for r in result.results)
+    assert len(result) == 2
+    assert all(r.success for r in result)
 
 
 # ---------------------------------------------------------------------------
diff --git a/isaaclab_arena/tests/test_object_placer_init.py b/isaaclab_arena/tests/test_object_placer_init.py
index e26883e1e..3510a03c5 100644
--- a/isaaclab_arena/tests/test_object_placer_init.py
+++ b/isaaclab_arena/tests/test_object_placer_init.py
@@ -222,7 +222,8 @@ def _run():
             bounding_box=AxisAlignedBoundingBox(min_point=(0.0, 0.0, 0.0), max_point=(0.2, 0.2, 0.2)),
         )
         box.add_relation(On(desk, clearance_m=0.01))
-        return ObjectPlacer(params=params).place([desk, box])
+        (result,) = ObjectPlacer(params=params).place([desk, box])
+        return result
 
     result1 = _run()
     result2 = _run()
diff --git a/isaaclab_arena/tests/test_object_placer_reproducibility.py b/isaaclab_arena/tests/test_object_placer_reproducibility.py
index 690018e12..10122fb0f 100644
--- a/isaaclab_arena/tests/test_object_placer_reproducibility.py
+++ b/isaaclab_arena/tests/test_object_placer_reproducibility.py
@@ -12,7 +12,7 @@
 from isaaclab_arena.assets.dummy_object import DummyObject
 from isaaclab_arena.relations.object_placer import ObjectPlacer
 from isaaclab_arena.relations.object_placer_params import ObjectPlacerParams
-from isaaclab_arena.relations.placement_result import MultiEnvPlacementResult, PlacementResult
+from isaaclab_arena.relations.placement_result import PlacementResult
 from isaaclab_arena.relations.pooled_object_placer import PooledObjectPlacer
 from isaaclab_arena.relations.relation_solver import RelationSolver
 from isaaclab_arena.relations.relation_solver_params import RelationSolverParams
@@ -97,13 +97,13 @@ def test_object_placer_same_seed_produces_identical_result():
     desk1, box1_run1, box2_run1 = _create_test_objects()
     objects1 = [desk1, box1_run1, box2_run1]
     placer1 = ObjectPlacer(params=ObjectPlacerParams(placement_seed=seed, solver_params=solver_params))
-    result1 = placer1.place(objects=objects1)
+    (result1,) = placer1.place(objects=objects1)
 
     # Run 2
     desk2, box1_run2, box2_run2 = _create_test_objects()
     objects2 = [desk2, box1_run2, box2_run2]
     placer2 = ObjectPlacer(params=ObjectPlacerParams(placement_seed=seed, solver_params=solver_params))
-    result2 = placer2.place(objects=objects2)
+    (result2,) = placer2.place(objects=objects2)
 
     # Compare by name
     for obj1, obj2 in zip(objects1, objects2):
@@ -121,13 +121,13 @@ def test_object_placer_different_seeds_produce_different_results():
     desk1, box1_run1, box2_run1 = _create_test_objects()
     objects1 = [desk1, box1_run1, box2_run1]
     placer1 = ObjectPlacer(params=ObjectPlacerParams(placement_seed=42, solver_params=solver_params))
-    result1 = placer1.place(objects=objects1)
+    (result1,) = placer1.place(objects=objects1)
 
     # Run 2 with seed 123
     desk2, box1_run2, box2_run2 = _create_test_objects()
     objects2 = [desk2, box1_run2, box2_run2]
     placer2 = ObjectPlacer(params=ObjectPlacerParams(placement_seed=123, solver_params=solver_params))
-    result2 = placer2.place(objects=objects2)
+    (result2,) = placer2.place(objects=objects2)
 
     # Check that at least one non-anchor position differs
     any_different = False
@@ -142,7 +142,7 @@ def test_object_placer_different_seeds_produce_different_results():
 
 
 def test_object_placer_multi_env_returns_multi_env_result():
-    """Test that ObjectPlacer.place with num_envs>1 returns MultiEnvPlacementResult."""
+    """place() with num_envs>1 returns one PlacementResult per env."""
     num_envs = 4
     solver_params = RelationSolverParams(max_iters=200, convergence_threshold=1e-3)
     desk, box1, box2 = _create_test_objects()
@@ -150,9 +150,8 @@ def test_object_placer_multi_env_returns_multi_env_result():
     placer = ObjectPlacer(params=ObjectPlacerParams(placement_seed=42, solver_params=solver_params))
     result = placer.place(objects, num_envs=num_envs)
 
-    assert isinstance(result, MultiEnvPlacementResult)
-    assert len(result.results) == num_envs
-    for r in result.results:
+    assert len(result) == num_envs
+    for r in result:
         assert isinstance(r, PlacementResult)
         assert box1 in r.positions
         assert box2 in r.positions
@@ -161,7 +160,7 @@ def test_object_placer_multi_env_returns_multi_env_result():
 
 
 def test_object_placer_multi_env_produces_different_positions():
-    """Test that multi-env placement produces different positions across environments."""
+    """Multi-env placement produces different positions across envs."""
     num_envs = 4
     solver_params = RelationSolverParams(max_iters=200, convergence_threshold=1e-3)
     desk, box1, box2 = _create_test_objects()
@@ -169,9 +168,8 @@ def test_object_placer_multi_env_produces_different_positions():
     placer = ObjectPlacer(params=ObjectPlacerParams(placement_seed=42, solver_params=solver_params))
     result = placer.place(objects, num_envs=num_envs)
 
-    assert isinstance(result, MultiEnvPlacementResult)
-    # At least one pair of envs should have different positions for a non-anchor object.
-    positions_box1 = [result.results[e].positions[box1] for e in range(num_envs)]
+    assert len(result) == num_envs
+    positions_box1 = [result[e].positions[box1] for e in range(num_envs)]
     any_different = any(positions_box1[i] != positions_box1[j] for i in range(num_envs) for j in range(i + 1, num_envs))
     assert any_different, "Multi-env placement should produce different positions across environments"
 
@@ -301,7 +299,7 @@ def test_random_yaw_init_applied_yaw_matches_selected_candidate():
     placer = ObjectPlacer(
         params=ObjectPlacerParams(placement_seed=11, solver_params=solver_params, random_yaw_init=True)
     )
-    result = placer.place([desk, box1, box2], num_envs=1)
+    (result,) = placer.place([desk, box1, box2], num_envs=1)
     for box in (box1, box2):
         applied = _yaw_rad_from_quat(box.get_initial_pose().rotation_xyzw)
         assert abs(wrap_angle_to_pi(applied - result.orientations[box])) < 1e-5
@@ -316,7 +314,7 @@ def test_random_yaw_init_composes_marker_yaw():
     placer = ObjectPlacer(
         params=ObjectPlacerParams(placement_seed=3, solver_params=solver_params, random_yaw_init=True)
     )
-    result = placer.place([desk, box1, box2], num_envs=1)
+    (result,) = placer.place([desk, box1, box2], num_envs=1)
     applied = _yaw_rad_from_quat(box1.get_initial_pose().rotation_xyzw)
     assert abs(wrap_angle_to_pi(applied - (marker_yaw + result.orientations[box1]))) < 1e-5
 
diff --git a/isaaclab_arena/tests/test_placement_events.py b/isaaclab_arena/tests/test_placement_events.py
index 1440d8070..aefa013bc 100644
--- a/isaaclab_arena/tests/test_placement_events.py
+++ b/isaaclab_arena/tests/test_placement_events.py
@@ -65,11 +65,11 @@ def test_successive_placements_without_seed_produce_different_layouts():
 
     desk1, box1_a, box2_a = _create_test_objects()
     placer_a = ObjectPlacer(params=params)
-    result_a = placer_a.place([desk1, box1_a, box2_a], num_envs=1)
+    (result_a,) = placer_a.place([desk1, box1_a, box2_a], num_envs=1)
 
     desk2, box1_b, box2_b = _create_test_objects()
     placer_b = ObjectPlacer(params=params)
-    result_b = placer_b.place([desk2, box1_b, box2_b], num_envs=1)
+    (result_b,) = placer_b.place([desk2, box1_b, box2_b], num_envs=1)
 
     any_different = False
     for obj_a, obj_b in zip([box1_a, box2_a], [box1_b, box2_b]):
@@ -84,7 +84,6 @@ def test_placement_without_seed_multi_env_gives_different_layouts():
 
     from isaaclab_arena.relations.object_placer import ObjectPlacer
     from isaaclab_arena.relations.object_placer_params import ObjectPlacerParams
-    from isaaclab_arena.relations.placement_result import MultiEnvPlacementResult
     from isaaclab_arena.relations.relation_solver_params import RelationSolverParams
 
     num_envs = 4
@@ -99,8 +98,8 @@ def test_placement_without_seed_multi_env_gives_different_layouts():
     placer = ObjectPlacer(params=params)
     result = placer.place([desk, box1, box2], num_envs=num_envs)
 
-    assert isinstance(result, MultiEnvPlacementResult)
-    positions_box1 = [result.results[env_idx].positions[box1] for env_idx in range(num_envs)]
+    assert len(result) == num_envs
+    positions_box1 = [result[env_idx].positions[box1] for env_idx in range(num_envs)]
     any_different = any(positions_box1[i] != positions_box1[j] for i in range(num_envs) for j in range(i + 1, num_envs))
     assert any_different, "Unseeded multi-env placement should produce different positions across environments"
 
@@ -121,11 +120,11 @@ def test_successive_seeded_placements_produce_same_layout():
 
     desk1, box1_a, box2_a = _create_test_objects()
     placer_a = ObjectPlacer(params=params)
-    result_a = placer_a.place([desk1, box1_a, box2_a], num_envs=1)
+    (result_a,) = placer_a.place([desk1, box1_a, box2_a], num_envs=1)
 
     desk2, box1_b, box2_b = _create_test_objects()
     placer_b = ObjectPlacer(params=params)
-    result_b = placer_b.place([desk2, box1_b, box2_b], num_envs=1)
+    (result_b,) = placer_b.place([desk2, box1_b, box2_b], num_envs=1)
 
     for obj_a, obj_b in zip([box1_a, box2_a], [box1_b, box2_b]):
         assert (
diff --git a/isaaclab_arena/tests/test_placement_result.py b/isaaclab_arena/tests/test_placement_result.py
new file mode 100644
index 000000000..3b17f8354
--- /dev/null
+++ b/isaaclab_arena/tests/test_placement_result.py
@@ -0,0 +1,48 @@
+# Copyright (c) 2025-2026, The Isaac Lab Arena Project Developers (https://github.com/isaac-sim/IsaacLab-Arena/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: Apache-2.0
+
+"""Unit tests for PlacementResult."""
+
+from isaaclab_arena.relations.placement_result import PlacementResult
+from isaaclab_arena.relations.placement_validation import PlacementCheck, PlacementValidationResults
+
+
+def _make_result(
+    *,
+    required: dict[PlacementCheck, bool] | None = None,
+    optional: dict[PlacementCheck, bool] | None = None,
+    final_loss: float = 0.0,
+) -> PlacementResult:
+    required = required or {PlacementCheck.NO_OVERLAP: True}
+    all_checks = dict(required)
+    if optional:
+        all_checks.update(optional)
+    return PlacementResult(
+        validation_results=PlacementValidationResults(
+            validation_results=all_checks,
+            required_checks=set(required),
+        ),
+        positions={},
+        final_loss=final_loss,
+        attempts=1,
+    )
+
+
+def test_success_true_when_all_required_pass():
+    result = _make_result(required={PlacementCheck.NO_OVERLAP: True, PlacementCheck.ON_RELATION: True})
+    assert result.success is True
+
+
+def test_success_false_when_required_fails():
+    result = _make_result(required={PlacementCheck.NO_OVERLAP: False})
+    assert result.success is False
+
+
+def test_success_ignores_failed_optional_check():
+    result = _make_result(
+        required={PlacementCheck.NO_OVERLAP: True},
+        optional={PlacementCheck.PHYSICS_SETTLED: False},
+    )
+    assert result.success is True
diff --git a/isaaclab_arena_examples/relations/dummy_object_placer_notebook.py b/isaaclab_arena_examples/relations/dummy_object_placer_notebook.py
index d67969157..66826f5a0 100644
--- a/isaaclab_arena_examples/relations/dummy_object_placer_notebook.py
+++ b/isaaclab_arena_examples/relations/dummy_object_placer_notebook.py
@@ -80,7 +80,7 @@ def run_dummy_object_placer_demo():
 
     # Place objects using ObjectPlacer (anchor is auto-detected via IsAnchor relation)
     placer = ObjectPlacer(params=ObjectPlacerParams())
-    result = placer.place(objects=all_objects)
+    (result,) = placer.place(objects=all_objects)
 
     # Visualization
     visualizer = RelationSolverVisualizer(
@@ -142,7 +142,7 @@ def run_dummy_multi_anchor_demo():
 
     # Place objects (verbose=True shows anchors and optimizable objects)
     placer = ObjectPlacer(params=ObjectPlacerParams(verbose=True))
-    result = placer.place(objects=all_objects)
+    (result,) = placer.place(objects=all_objects)
 
     print("\nFinal positions:")
     for obj, pos in result.positions.items():
diff --git a/isaaclab_arena_examples/relations/isaac_sim_no_collision_notebook.py b/isaaclab_arena_examples/relations/isaac_sim_no_collision_notebook.py
index 2fc6a11a7..74006d7cc 100644
--- a/isaaclab_arena_examples/relations/isaac_sim_no_collision_notebook.py
+++ b/isaaclab_arena_examples/relations/isaac_sim_no_collision_notebook.py
@@ -129,7 +129,7 @@ def apply_overlapping_pose_then_solve_and_display():
             env.unwrapped.sim.render()
         # Run the solver and apply solved poses to the sim.
         placer = ObjectPlacer()
-        result = placer.place(objects=objects_with_relations)
+        (result,) = placer.place(objects=objects_with_relations)
         for obj in placeable_objects:
             if obj.name not in env.unwrapped.scene.rigid_objects:
                 print(