diff --git a/README.md b/README.md
index 6103240..53a0e31 100644
--- a/README.md
+++ b/README.md
@@ -1006,9 +1006,6 @@ T step(T edge, T x);
 
 template < floating_point T >
 T smoothstep(T edge0, T edge1, T x);
-
-template < floating_point T >
-T fma(T x, T y, T z);
 ```
 
 #### Vector
@@ -1112,9 +1109,6 @@ vec<T, Size> smoothstep(T edge0, T edge1, const vec<T, Size>& xs);
 
 template < typename T, size_t Size >
 vec<T, Size> smoothstep(const vec<T, Size>& edges0, const vec<T, Size>& edges1, const vec<T, Size>& xs);
-
-template < typename T, size_t Size >
-vec<T, Size> fma(const vec<T, Size>& as, const vec<T, Size>& bs, const vec<T, Size>& cs);
 ```
 
 #### Quaternion
diff --git a/headers/vmath.hpp/vmath_fun.hpp b/headers/vmath.hpp/vmath_fun.hpp
index e87e262..8da26fc 100644
--- a/headers/vmath.hpp/vmath_fun.hpp
+++ b/headers/vmath.hpp/vmath_fun.hpp
@@ -138,12 +138,6 @@ namespace vmath_hpp
         const T t = clamp((x - edge0) * rcp(edge1 - edge0), T{0}, T{1});
         return t * t * (T{3} - T{2} * t);
     }
-
-    template < typename T >
-    [[nodiscard]] std::enable_if_t<std::is_floating_point_v<T>, T>
-    fma(T x, T y, T z) noexcept {
-        return std::fma(x, y, z);
-    }
 }
 
 //
diff --git a/headers/vmath.hpp/vmath_vec_fun.hpp b/headers/vmath.hpp/vmath_vec_fun.hpp
index 9cf110e..69c3c6e 100644
--- a/headers/vmath.hpp/vmath_vec_fun.hpp
+++ b/headers/vmath.hpp/vmath_vec_fun.hpp
@@ -632,11 +632,6 @@ namespace vmath_hpp
     [[nodiscard]] constexpr vec<T, Size> smoothstep(const vec<T, Size>& edges0, const vec<T, Size>& edges1, const vec<T, Size>& xs) {
         return map_join([](T edge0, T edge1, T x) { return smoothstep(edge0, edge1, x); }, edges0, edges1, xs);
     }
-
-    template < typename T, std::size_t Size >
-    [[nodiscard]] constexpr vec<T, Size> fma(const vec<T, Size>& as, const vec<T, Size>& bs, const vec<T, Size>& cs) {
-        return map_join([](T a, T b, T c) { return fma(a, b, c); }, as, bs, cs);
-    }
 }
 
 //
diff --git a/untests/vmath_fix_tests.cpp b/untests/vmath_fix_tests.cpp
index 1d0d4ad..1af899d 100644
--- a/untests/vmath_fix_tests.cpp
+++ b/untests/vmath_fix_tests.cpp
@@ -47,8 +47,6 @@ namespace
         constexpr friend fix step(const fix& e, const fix& l) { using vmath_hpp::step; return fix{step(e.underlying(), l.underlying())}; }
         constexpr friend fix smoothstep(const fix& e0, const fix& e1, const fix& l) { using vmath_hpp::smoothstep; return fix{smoothstep(e0.underlying(), e1.underlying(), l.underlying())}; }
 
-        constexpr friend fix fma(const fix& x, const fix& y, const fix& z) { using vmath_hpp::fma; return fix{fma(x.underlying(), y.underlying(), z.underlying())}; }
-
         //
 
         constexpr friend fix<bool> any(const fix& l) { using vmath_hpp::any; return fix<bool>{any(l.underlying())}; }
@@ -199,8 +197,6 @@ namespace vmath_hpp
     template fix2f step(const fix2f&, const fix2f&);
     template fix2f smoothstep(fix<float>, fix<float>, const fix2f&);
     template fix2f smoothstep(const fix2f&, const fix2f&, const fix2f&);
-
-    template fix2f fma(const fix2f&, const fix2f&, const fix2f&);
 }
 
 namespace vmath_hpp
diff --git a/untests/vmath_fun_tests.cpp b/untests/vmath_fun_tests.cpp
index 4817f09..9903781 100644
--- a/untests/vmath_fun_tests.cpp
+++ b/untests/vmath_fun_tests.cpp
@@ -105,8 +105,6 @@ TEST_CASE("vmath/fun") {
         STATIC_CHECK(step(0.5f, 0.4f) == uapprox(0.f));
         STATIC_CHECK(step(0.5f, 0.6f) == uapprox(1.f));
         STATIC_CHECK(smoothstep(0.f, 1.f, 0.1f) == uapprox(0.028f));
-
-        CHECK(fma(2.f, 3.f, 4.f) == uapprox(10.f));
     }
 
     SUBCASE("Geometric Functions") {
diff --git a/untests/vmath_vec_fun_tests.cpp b/untests/vmath_vec_fun_tests.cpp
index 31b4915..262f1ec 100644
--- a/untests/vmath_vec_fun_tests.cpp
+++ b/untests/vmath_vec_fun_tests.cpp
@@ -249,8 +249,6 @@ TEST_CASE("vmath/vec_fun") {
 
         STATIC_CHECK(smoothstep(0.f, 1.f, float2(0.1f)) == uapprox2(0.028f));
         STATIC_CHECK(smoothstep(float2(0.f), float2(1.f), float2(0.1f)) == uapprox2(0.028f));
-
-        CHECK_FALSE(fma(float2(2.f), float2(3.f), float2(4.f)).x == uapprox(12.f));
     }
 
     SUBCASE("Geometric Functions") {