diff --git a/include/fmt/format-inl.h b/include/fmt/format-inl.h
index 5f07a465..d812c3e1 100644
--- a/include/fmt/format-inl.h
+++ b/include/fmt/format-inl.h
@@ -1038,7 +1038,7 @@ void fallback_format(Double d, buffer<char>& buf, int& exp10) {
 // if T is a IEEE754 binary32 or binary64 and snprintf otherwise.
 template <typename T>
 int format_float(T value, int precision, float_specs specs, buffer<char>& buf) {
-  static_assert(!std::is_same<T, float>(), "");
+  static_assert(!std::is_same<T, float>::value, "");
   FMT_ASSERT(value >= 0, "value is negative");
 
   const bool fixed = specs.format == float_format::fixed;
@@ -1113,7 +1113,7 @@ int snprintf_float(T value, int precision, float_specs specs,
                    buffer<char>& buf) {
   // Buffer capacity must be non-zero, otherwise MSVC's vsnprintf_s will fail.
   FMT_ASSERT(buf.capacity() > buf.size(), "empty buffer");
-  static_assert(!std::is_same<T, float>(), "");
+  static_assert(!std::is_same<T, float>::value, "");
 
   // Subtract 1 to account for the difference in precision since we use %e for
   // both general and exponent format.