diff --git a/core/math/math_2d.cpp b/core/math/math_2d.cpp
index d2e4101999d6..3767d298a187 100644
--- a/core/math/math_2d.cpp
+++ b/core/math/math_2d.cpp
@@ -98,11 +98,6 @@ real_t Vector2::cross(const Vector2 &p_other) const {
 	return x * p_other.y - y * p_other.x;
 }
 
-Vector2 Vector2::cross(real_t p_other) const {
-
-	return Vector2(p_other * y, -p_other * x);
-}
-
 Vector2 Vector2::floor() const {
 
 	return Vector2(Math::floor(x), Math::floor(y));
diff --git a/core/math/math_2d.h b/core/math/math_2d.h
index 8928349a448b..02d921b67e27 100644
--- a/core/math/math_2d.h
+++ b/core/math/math_2d.h
@@ -104,7 +104,6 @@ struct Vector2 {
 
 	real_t dot(const Vector2 &p_other) const;
 	real_t cross(const Vector2 &p_other) const;
-	Vector2 cross(real_t p_other) const;
 	Vector2 project(const Vector2 &p_vec) const;
 
 	Vector2 plane_project(real_t p_d, const Vector2 &p_vec) const;
diff --git a/core/variant_call.cpp b/core/variant_call.cpp
index d1d45f0e7c2d..cda7dccf0cb2 100644
--- a/core/variant_call.cpp
+++ b/core/variant_call.cpp
@@ -347,7 +347,7 @@ struct _VariantCall {
 	VCALL_LOCALMEM1R(Vector2, bounce);
 	VCALL_LOCALMEM1R(Vector2, reflect);
 	VCALL_LOCALMEM0R(Vector2, angle);
-	//VCALL_LOCALMEM1R(Vector2,cross);
+	VCALL_LOCALMEM1R(Vector2, cross);
 	VCALL_LOCALMEM0R(Vector2, abs);
 	VCALL_LOCALMEM1R(Vector2, clamped);
 
@@ -1517,7 +1517,7 @@ void register_variant_methods() {
 	ADDFUNC1R(VECTOR2, VECTOR2, Vector2, slide, VECTOR2, "n", varray());
 	ADDFUNC1R(VECTOR2, VECTOR2, Vector2, bounce, VECTOR2, "n", varray());
 	ADDFUNC1R(VECTOR2, VECTOR2, Vector2, reflect, VECTOR2, "n", varray());
-	//ADDFUNC1R(VECTOR2,REAL,Vector2,cross,VECTOR2,"with",varray());
+	ADDFUNC1R(VECTOR2, REAL, Vector2, cross, VECTOR2, "with", varray());
 	ADDFUNC0R(VECTOR2, VECTOR2, Vector2, abs, varray());
 	ADDFUNC1R(VECTOR2, VECTOR2, Vector2, clamped, REAL, "length", varray());
 
diff --git a/doc/classes/Vector2.xml b/doc/classes/Vector2.xml
index 9df2eb143b78..b293ec3368b0 100644
--- a/doc/classes/Vector2.xml
+++ b/doc/classes/Vector2.xml
@@ -81,6 +81,15 @@
 				Returns the vector with a maximum length.
 			</description>
 		</method>
+		<method name="cross">
+			<return type="float">
+			</return>
+			<argument index="0" name="b" type="Vector2">
+			</argument>
+			<description>
+				Returns the 2-dimensional analog of the cross product with [code]b[/code].
+			</description>
+		</method>
 		<method name="cubic_interpolate">
 			<return type="Vector2">
 			</return>
diff --git a/servers/physics_2d/joints_2d_sw.cpp b/servers/physics_2d/joints_2d_sw.cpp
index 7fba8acebd89..d49c1b837671 100644
--- a/servers/physics_2d/joints_2d_sw.cpp
+++ b/servers/physics_2d/joints_2d_sw.cpp
@@ -146,14 +146,19 @@ bool PinJoint2DSW::setup(real_t p_step) {
 	return true;
 }
 
+inline Vector2 custom_cross(const Vector2 &p_vec, real_t p_other) {
+
+	return Vector2(p_other * p_vec.y, -p_other * p_vec.x);
+}
+
 void PinJoint2DSW::solve(real_t p_step) {
 
 	// compute relative velocity
-	Vector2 vA = A->get_linear_velocity() - rA.cross(A->get_angular_velocity());
+	Vector2 vA = A->get_linear_velocity() - custom_cross(rA, A->get_angular_velocity());
 
 	Vector2 rel_vel;
 	if (B)
-		rel_vel = B->get_linear_velocity() - rB.cross(B->get_angular_velocity()) - vA;
+		rel_vel = B->get_linear_velocity() - custom_cross(rB, B->get_angular_velocity()) - vA;
 	else
 		rel_vel = -vA;