diff --git a/test/test_transforms.py b/test/test_transforms.py
index b3c82334d14..0562adb2a90 100644
--- a/test/test_transforms.py
+++ b/test/test_transforms.py
@@ -180,6 +180,14 @@ def test_randomperspective(self):
                                torch.nn.functional.mse_loss(tr_img2, F.to_tensor(img)))
 
     def test_randomperspective_fill(self):
+
+        # assert fill being either a Sequence or a Number
+        with self.assertRaises(TypeError):
+            transforms.RandomPerspective(fill={})
+
+        t = transforms.RandomPerspective(fill=None)
+        self.assertTrue(t.fill == 0)
+
         height = 100
         width = 100
         img = torch.ones(3, height, width)
@@ -1531,6 +1539,13 @@ def test_random_rotation(self):
             transforms.RandomRotation([-0.7])
             transforms.RandomRotation([-0.7, 0, 0.7])
 
+        # assert fill being either a Sequence or a Number
+        with self.assertRaises(TypeError):
+            transforms.RandomRotation(0, fill={})
+
+        t = transforms.RandomRotation(0, fill=None)
+        self.assertTrue(t.fill == 0)
+
         t = transforms.RandomRotation(10)
         angle = t.get_params(t.degrees)
         self.assertTrue(angle > -10 and angle < 10)
@@ -1573,6 +1588,13 @@ def test_random_affine(self):
             transforms.RandomAffine([-90, 90], translate=[0.2, 0.2], scale=[0.5, 0.5], shear=[-10, 0, 10])
             transforms.RandomAffine([-90, 90], translate=[0.2, 0.2], scale=[0.5, 0.5], shear=[-10, 0, 10, 0, 10])
 
+        # assert fill being either a Sequence or a Number
+        with self.assertRaises(TypeError):
+            transforms.RandomAffine(0, fill={})
+
+        t = transforms.RandomAffine(0, fill=None)
+        self.assertTrue(t.fill == 0)
+
         x = np.zeros((100, 100, 3), dtype=np.uint8)
         img = F.to_pil_image(x)
 
diff --git a/torchvision/transforms/transforms.py b/torchvision/transforms/transforms.py
index bf847c8fc75..ff12a070571 100644
--- a/torchvision/transforms/transforms.py
+++ b/torchvision/transforms/transforms.py
@@ -673,8 +673,8 @@ class RandomPerspective(torch.nn.Module):
             :class:`torchvision.transforms.InterpolationMode`. Default is ``InterpolationMode.BILINEAR``.
             If input is Tensor, only ``InterpolationMode.NEAREST``, ``InterpolationMode.BILINEAR`` are supported.
             For backward compatibility integer values (e.g. ``PIL.Image.NEAREST``) are still acceptable.
-        fill (sequence or number, optional): Pixel fill value for the area outside the transformed
-            image. If given a number, the value is used for all bands respectively.
+        fill (sequence or number): Pixel fill value for the area outside the transformed
+            image. Default is ``0``. If given a number, the value is used for all bands respectively.
             If input is PIL Image, the options is only available for ``Pillow>=5.0.0``.
     """
 
@@ -692,6 +692,12 @@ def __init__(self, distortion_scale=0.5, p=0.5, interpolation=InterpolationMode.
 
         self.interpolation = interpolation
         self.distortion_scale = distortion_scale
+
+        if fill is None:
+            fill = 0
+        elif not isinstance(fill, (Sequence, numbers.Number)):
+            raise TypeError("Fill should be either a sequence or a number.")
+
         self.fill = fill
 
     def forward(self, img):
@@ -1175,8 +1181,8 @@ class RandomRotation(torch.nn.Module):
             Note that the expand flag assumes rotation around the center and no translation.
         center (sequence, optional): Optional center of rotation, (x, y). Origin is the upper left corner.
             Default is the center of the image.
-        fill (sequence or number, optional): Pixel fill value for the area outside the rotated
-            image. If given a number, the value is used for all bands respectively.
+        fill (sequence or number): Pixel fill value for the area outside the rotated
+            image. Default is ``0``. If given a number, the value is used for all bands respectively.
             If input is PIL Image, the options is only available for ``Pillow>=5.2.0``.
         resample (int, optional): deprecated argument and will be removed since v0.10.0.
             Please use the ``interpolation`` parameter instead.
@@ -1186,7 +1192,7 @@ class RandomRotation(torch.nn.Module):
     """
 
     def __init__(
-        self, degrees, interpolation=InterpolationMode.NEAREST, expand=False, center=None, fill=None, resample=None
+        self, degrees, interpolation=InterpolationMode.NEAREST, expand=False, center=None, fill=0, resample=None
     ):
         super().__init__()
         if resample is not None:
@@ -1212,6 +1218,12 @@ def __init__(
 
         self.resample = self.interpolation = interpolation
         self.expand = expand
+
+        if fill is None:
+            fill = 0
+        elif not isinstance(fill, (Sequence, numbers.Number)):
+            raise TypeError("Fill should be either a sequence or a number.")
+
         self.fill = fill
 
     @staticmethod
@@ -1280,8 +1292,8 @@ class RandomAffine(torch.nn.Module):
             :class:`torchvision.transforms.InterpolationMode`. Default is ``InterpolationMode.NEAREST``.
             If input is Tensor, only ``InterpolationMode.NEAREST``, ``InterpolationMode.BILINEAR`` are supported.
             For backward compatibility integer values (e.g. ``PIL.Image.NEAREST``) are still acceptable.
-        fill (sequence or number, optional): Pixel fill value for the area outside the transformed
-            image. If given a number, the value is used for all bands respectively.
+        fill (sequence or number): Pixel fill value for the area outside the transformed
+            image. Default is ``0``. If given a number, the value is used for all bands respectively.
             If input is PIL Image, the options is only available for ``Pillow>=5.0.0``.
         fillcolor (sequence or number, optional): deprecated argument and will be removed since v0.10.0.
             Please use the ``fill`` parameter instead.
@@ -1339,6 +1351,12 @@ def __init__(
             self.shear = shear
 
         self.resample = self.interpolation = interpolation
+
+        if fill is None:
+            fill = 0
+        elif not isinstance(fill, (Sequence, numbers.Number)):
+            raise TypeError("Fill should be either a sequence or a number.")
+
         self.fillcolor = self.fill = fill
 
     @staticmethod