diff --git a/numpy/ma/core.pyi b/numpy/ma/core.pyi
index 3fad41a1434e..ca9826c7d083 100644
--- a/numpy/ma/core.pyi
+++ b/numpy/ma/core.pyi
@@ -1200,7 +1200,12 @@ class MaskedArray(ndarray[_ShapeT_co, _DTypeT_co]):
         out: _ArrayT,
     ) -> _ArrayT: ...
 
-    def dot(self, b, out=..., strict=...): ...
+    # This differs from `ndarray.dot`, in that 1D dot 1D returns a 0D array.
+    @overload
+    def dot(self, b: ArrayLike, out: None = ..., strict: bool = ...) -> _MaskedArray[Any]: ...
+    @overload
+    def dot(self, b: ArrayLike, out: _ArrayT, strict: bool = ...) -> _ArrayT: ...
+
     def sum(self, axis=..., dtype=..., out=..., keepdims=...): ...
 
     # Keep in sync with `ndarray.cumsum`
@@ -1223,7 +1228,14 @@ class MaskedArray(ndarray[_ShapeT_co, _DTypeT_co]):
     def cumprod(self, /, axis: SupportsIndex | None = None, dtype: DTypeLike | None = None, *, out: _ArrayT) -> _ArrayT: ...
 
     def mean(self, axis=..., dtype=..., out=..., keepdims=...): ...
-    def anom(self, axis=..., dtype=...): ...
+
+    @overload
+    def anom(self, axis: SupportsIndex | None = None, dtype: None = None) -> Self: ...
+    @overload
+    def anom(self, axis: SupportsIndex | None = None, *, dtype: DTypeLike) -> MaskedArray[_ShapeT_co, dtype]: ...
+    @overload
+    def anom(self, axis: SupportsIndex | None, dtype: DTypeLike) -> MaskedArray[_ShapeT_co, dtype]: ...
+
     def var(self, axis=..., dtype=..., out=..., ddof=..., keepdims=...): ...
     def std(self, axis=..., dtype=..., out=..., ddof=..., keepdims=...): ...
 
diff --git a/numpy/typing/tests/data/reveal/ma.pyi b/numpy/typing/tests/data/reveal/ma.pyi
index 9e8ab98ec4ce..c26dca0ed9ec 100644
--- a/numpy/typing/tests/data/reveal/ma.pyi
+++ b/numpy/typing/tests/data/reveal/ma.pyi
@@ -373,6 +373,11 @@ assert_type(MAR_b.shrink_mask(), MaskedArray[np.bool_])
 assert_type(MAR_i8.hardmask, bool)
 assert_type(MAR_i8.sharedmask, bool)
 
+assert_type(MAR_2d_f4.anom(), np.ma.MaskedArray[tuple[int, int], np.dtype[np.float32]])
+assert_type(MAR_2d_f4.anom(axis=0, dtype=np.float16), np.ma.MaskedArray[tuple[int, int], np.dtype])
+assert_type(MAR_2d_f4.anom(0, np.float16), np.ma.MaskedArray[tuple[int, int], np.dtype])
+assert_type(MAR_2d_f4.anom(0, 'float16'), np.ma.MaskedArray[tuple[int, int], np.dtype])
+
 assert_type(MAR_b.transpose(), MaskedArray[np.bool])
 assert_type(MAR_2d_f4.transpose(), np.ma.MaskedArray[tuple[int, int], np.dtype[np.float32]])
 assert_type(MAR_2d_f4.transpose(1, 0), np.ma.MaskedArray[tuple[int, int], np.dtype[np.float32]])
@@ -380,6 +385,10 @@ assert_type(MAR_2d_f4.transpose((1, 0)), np.ma.MaskedArray[tuple[int, int], np.d
 assert_type(MAR_b.T, MaskedArray[np.bool])
 assert_type(MAR_2d_f4.T, np.ma.MaskedArray[tuple[int, int], np.dtype[np.float32]])
 
+assert_type(MAR_2d_f4.dot(1), MaskedArray[Any])
+assert_type(MAR_2d_f4.dot([1]), MaskedArray[Any])
+assert_type(MAR_2d_f4.dot(1, out=MAR_subclass), MaskedArraySubclass)
+
 assert_type(MAR_2d_f4.nonzero(), tuple[_Array1D[np.intp], ...])
 assert_type(MAR_2d_f4.nonzero()[0], _Array1D[np.intp])