diff --git a/pandas/core/arrays/masked.py b/pandas/core/arrays/masked.py
index 1797f1aff4235..568f3484e78e4 100644
--- a/pandas/core/arrays/masked.py
+++ b/pandas/core/arrays/masked.py
@@ -47,6 +47,7 @@
 )
 from pandas.core.dtypes.inference import is_array_like
 from pandas.core.dtypes.missing import (
+    array_equivalent,
     isna,
     notna,
 )
@@ -636,11 +637,12 @@ def equals(self, other) -> bool:
 
         # GH#44382 if e.g. self[1] is np.nan and other[1] is pd.NA, we are NOT
         #  equal.
-        return np.array_equal(self._mask, other._mask) and np.array_equal(
-            self._data[~self._mask],
-            other._data[~other._mask],
-            equal_nan=True,
-        )
+        if not np.array_equal(self._mask, other._mask):
+            return False
+
+        left = self._data[~self._mask]
+        right = other._data[~other._mask]
+        return array_equivalent(left, right, dtype_equal=True)
 
     def _reduce(self, name: str, *, skipna: bool = True, **kwargs):
         if name in {"any", "all"}:
diff --git a/pandas/core/dtypes/missing.py b/pandas/core/dtypes/missing.py
index c457b52cf4b0e..eea3fa37b7435 100644
--- a/pandas/core/dtypes/missing.py
+++ b/pandas/core/dtypes/missing.py
@@ -475,8 +475,8 @@ def array_equivalent(
     return np.array_equal(left, right)
 
 
-def _array_equivalent_float(left, right):
-    return ((left == right) | (np.isnan(left) & np.isnan(right))).all()
+def _array_equivalent_float(left, right) -> bool:
+    return bool(((left == right) | (np.isnan(left) & np.isnan(right))).all())
 
 
 def _array_equivalent_datetimelike(left, right):