diff --git a/include/boost/multiprecision/mpfr.hpp b/include/boost/multiprecision/mpfr.hpp
index b774e60e..a7bcb9a8 100644
--- a/include/boost/multiprecision/mpfr.hpp
+++ b/include/boost/multiprecision/mpfr.hpp
@@ -1105,8 +1105,8 @@ inline void eval_divide(mpfr_float_backend<digits10, AllocationType>& result, lo
 //
 // Specialised 3 arg versions of the basic operators:
 //
-template <unsigned D1, unsigned D2, mpfr_allocation_type A1, mpfr_allocation_type A2, unsigned D3>
-inline void eval_add(mpfr_float_backend<D1, A1>& a, const mpfr_float_backend<D2, A2>& x, const mpfr_float_backend<D3>& y)
+template <unsigned D1, unsigned D2, mpfr_allocation_type A1, mpfr_allocation_type A2, unsigned D3, mpfr_allocation_type A3>
+inline void eval_add(mpfr_float_backend<D1, A1>& a, const mpfr_float_backend<D2, A2>& x, const mpfr_float_backend<D3, A3>& y)
 {
    mpfr_add(a.data(), x.data(), y.data(), GMP_RNDN);
 }
@@ -1139,8 +1139,8 @@ inline void eval_add(mpfr_float_backend<D1, A1>& a, long x, const mpfr_float_bac
    else
       mpfr_add_ui(a.data(), y.data(), x, GMP_RNDN);
 }
-template <unsigned D1, unsigned D2, mpfr_allocation_type A1, mpfr_allocation_type A2, unsigned D3>
-inline void eval_subtract(mpfr_float_backend<D1, A1>& a, const mpfr_float_backend<D2, A2>& x, const mpfr_float_backend<D3>& y)
+template <unsigned D1, unsigned D2, mpfr_allocation_type A1, mpfr_allocation_type A2, unsigned D3, mpfr_allocation_type A3>
+inline void eval_subtract(mpfr_float_backend<D1, A1>& a, const mpfr_float_backend<D2, A2>& x, const mpfr_float_backend<D3, A3>& y)
 {
    mpfr_sub(a.data(), x.data(), y.data(), GMP_RNDN);
 }
@@ -1174,8 +1174,8 @@ inline void eval_subtract(mpfr_float_backend<D1, A1>& a, long x, const mpfr_floa
       mpfr_ui_sub(a.data(), x, y.data(), GMP_RNDN);
 }
 
-template <unsigned D1, unsigned D2, mpfr_allocation_type A1, mpfr_allocation_type A2, unsigned D3>
-inline void eval_multiply(mpfr_float_backend<D1, A1>& a, const mpfr_float_backend<D2, A2>& x, const mpfr_float_backend<D3>& y)
+template <unsigned D1, unsigned D2, mpfr_allocation_type A1, mpfr_allocation_type A2, unsigned D3, mpfr_allocation_type A3>
+inline void eval_multiply(mpfr_float_backend<D1, A1>& a, const mpfr_float_backend<D2, A2>& x, const mpfr_float_backend<D3, A3>& y)
 {
    if ((void*)&x == (void*)&y)
       mpfr_sqr(a.data(), x.data(), GMP_RNDN);
@@ -1215,8 +1215,8 @@ inline void eval_multiply(mpfr_float_backend<D1, A1>& a, long x, const mpfr_floa
       mpfr_mul_ui(a.data(), y.data(), x, GMP_RNDN);
 }
 
-template <unsigned D1, unsigned D2, mpfr_allocation_type A1, mpfr_allocation_type A2, unsigned D3>
-inline void eval_divide(mpfr_float_backend<D1, A1>& a, const mpfr_float_backend<D2, A2>& x, const mpfr_float_backend<D3>& y)
+template <unsigned D1, unsigned D2, mpfr_allocation_type A1, mpfr_allocation_type A2, unsigned D3, mpfr_allocation_type A3>
+inline void eval_divide(mpfr_float_backend<D1, A1>& a, const mpfr_float_backend<D2, A2>& x, const mpfr_float_backend<D3, A3>& y)
 {
    mpfr_div(a.data(), x.data(), y.data(), GMP_RNDN);
 }
@@ -1632,6 +1632,9 @@ template <>
 struct number_category<detail::canonical<mpfr_t, backends::mpfr_float_backend<0> >::type> : public mpl::int_<number_kind_floating_point>
 {};
 
+template <unsigned D, boost::multiprecision::mpfr_allocation_type A1, boost::multiprecision::mpfr_allocation_type A2>
+struct is_equivalent_number_type<backends::mpfr_float_backend<D, A1>, backends::mpfr_float_backend<D, A2> > : public mpl::true_ {};
+
 using boost::multiprecision::backends::mpfr_float_backend;
 
 typedef number<mpfr_float_backend<50> >   mpfr_float_50;
diff --git a/test/test_arithmetic_mpfr_50_mixed.cpp b/test/test_arithmetic_mpfr_50_mixed.cpp
new file mode 100644
index 00000000..5810e09b
--- /dev/null
+++ b/test/test_arithmetic_mpfr_50_mixed.cpp
@@ -0,0 +1,24 @@
+///////////////////////////////////////////////////////////////
+//  Copyright 2012 John Maddock. Distributed under the Boost
+//  Software License, Version 1.0. (See accompanying file
+//  LICENSE_1_0.txt or copy at https://www.boost.org/LICENSE_1_0.txt
+
+#ifdef _MSC_VER
+#define _SCL_SECURE_NO_WARNINGS
+#endif
+
+#include <boost/multiprecision/mpfr.hpp>
+#define TEST_MPFR
+#include "test_arithmetic.hpp"
+
+template <>
+struct related_type<boost::multiprecision::static_mpfr_float_50>
+{
+   typedef boost::multiprecision::mpfr_float_50 type;
+};
+
+int main()
+{
+   test<boost::multiprecision::static_mpfr_float_50>();
+   return boost::report_errors();
+}