diff --git a/doc/syntax.rst b/doc/syntax.rst
index ab8787df256c..f6f43afb3a8a 100644
--- a/doc/syntax.rst
+++ b/doc/syntax.rst
@@ -157,7 +157,8 @@ displayed after the decimal point for a floating-point value formatted with
 ``'f'`` and ``'F'``, or before and after the decimal point for a floating-point
 value formatted with ``'g'`` or ``'G'``.  For non-number types the field
 indicates the maximum field size - in other words, how many characters will be
-used from the field content. The *precision* is not allowed for integer values.
+used from the field content. The *precision* is not allowed for integer values
+or pointers.
 
 Finally, the *type* determines how the data should be presented.
 
diff --git a/format.cc b/format.cc
index 01380504a099..0c400aef4eaa 100644
--- a/format.cc
+++ b/format.cc
@@ -561,9 +561,13 @@ class fmt::internal::ArgFormatter :
     if (spec_.align_ == ALIGN_NUMERIC || spec_.flags_ != 0)
       FMT_THROW(FormatError("invalid format specifier for char"));
     typedef typename fmt::BasicWriter<Char>::CharPtr CharPtr;
+    Char fill = static_cast<Char>(spec_.fill());
+    if (spec_.precision_ == 0) {
+      std::fill_n(writer_.grow_buffer(spec_.width_), spec_.width_, fill);
+      return;
+    }
     CharPtr out = CharPtr();
     if (spec_.width_ > 1) {
-      Char fill = static_cast<Char>(spec_.fill());
       out = writer_.grow_buffer(spec_.width_);
       if (spec_.align_ == fmt::ALIGN_RIGHT) {
         std::fill_n(out, spec_.width_ - 1, fill);
@@ -615,6 +619,8 @@ void fmt::BasicWriter<Char>::write_str(
     if (*str_value)
       str_size = std::char_traits<StrChar>::length(str_value);
   }
+  if (spec.precision_ >= 0 && spec.precision_ < str_size)
+    str_size = spec.precision_;
   write_str(str_value, str_size, spec);
 }
 
@@ -1011,9 +1017,10 @@ const Char *fmt::BasicFormatter<Char>::format(
       } else {
         FMT_THROW(FormatError("missing precision specifier"));
       }
-      if (arg.type != Arg::DOUBLE && arg.type != Arg::LONG_DOUBLE) {
+      if (arg.type < Arg::LAST_INTEGER_TYPE || arg.type == Arg::POINTER) {
         FMT_THROW(FormatError(
-            "precision specifier requires floating-point argument"));
+            fmt::format("precision not allowed in {} format specifier",
+            arg.type == Arg::POINTER ? "pointer" : "integer")));
       }
     }
 
diff --git a/test/format-test.cc b/test/format-test.cc
index 6aa79cb8eecb..d1f4f074f23f 100644
--- a/test/format-test.cc
+++ b/test/format-test.cc
@@ -824,53 +824,42 @@ TEST(FormatterTest, Precision) {
       FormatError, "missing precision specifier");
 
   EXPECT_THROW_MSG(format("{0:.2", 0),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.2}", 42),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.2f}", 42),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.2}", 42u),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.2f}", 42u),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.2}", 42l),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.2f}", 42l),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.2}", 42ul),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.2f}", 42ul),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.2}", 42ll),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.2f}", 42ll),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.2}", 42ull),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.2f}", 42ull),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_EQ("1.2", format("{0:.2}", 1.2345));
   EXPECT_EQ("1.2", format("{0:.2}", 1.2345l));
 
   EXPECT_THROW_MSG(format("{0:.2}", reinterpret_cast<void*>(0xcafe)),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in pointer format specifier");
   EXPECT_THROW_MSG(format("{0:.2f}", reinterpret_cast<void*>(0xcafe)),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in pointer format specifier");
 
-  EXPECT_THROW_MSG(format("{0:.2}", 'x'),
-      FormatError, "precision specifier requires floating-point argument");
-  EXPECT_THROW_MSG(format("{0:.2f}", 'x'),
-      FormatError, "precision specifier requires floating-point argument");
-
-  EXPECT_THROW_MSG(format("{0:.2}", "str"),
-      FormatError, "precision specifier requires floating-point argument");
-  EXPECT_THROW_MSG(format("{0:.2f}", "str"),
-      FormatError, "precision specifier requires floating-point argument");
-
-  EXPECT_THROW_MSG(format("{0:.2}", TestString()),
-      FormatError, "precision specifier requires floating-point argument");
-  EXPECT_THROW_MSG(format("{0:.2f}", TestString()),
-      FormatError, "precision specifier requires floating-point argument");
+  EXPECT_EQ("   ", format("{0:3.0}", 'x'));
+  EXPECT_EQ("st", format("{0:.2}", "str"));
+  EXPECT_EQ("te", format("{0:.2}", TestString("test")));
 }
 
 TEST(FormatterTest, RuntimePrecision) {
@@ -893,7 +882,7 @@ TEST(FormatterTest, RuntimePrecision) {
   EXPECT_THROW_MSG(format("{0:.{x}}", 0),
       FormatError, "invalid format string");
   EXPECT_THROW_MSG(format("{0:.{1}", 0, 0),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.{1}}", 0),
       FormatError, "argument index out of range");
 
@@ -920,51 +909,40 @@ TEST(FormatterTest, RuntimePrecision) {
       FormatError, "precision is not integer");
 
   EXPECT_THROW_MSG(format("{0:.{1}}", 42, 2),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.{1}f}", 42, 2),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.{1}}", 42u, 2),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.{1}f}", 42u, 2),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.{1}}", 42l, 2),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.{1}f}", 42l, 2),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.{1}}", 42ul, 2),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.{1}f}", 42ul, 2),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.{1}}", 42ll, 2),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.{1}f}", 42ll, 2),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.{1}}", 42ull, 2),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_THROW_MSG(format("{0:.{1}f}", 42ull, 2),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in integer format specifier");
   EXPECT_EQ("1.2", format("{0:.{1}}", 1.2345, 2));
   EXPECT_EQ("1.2", format("{1:.{0}}", 2, 1.2345l));
 
   EXPECT_THROW_MSG(format("{0:.{1}}", reinterpret_cast<void*>(0xcafe), 2),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in pointer format specifier");
   EXPECT_THROW_MSG(format("{0:.{1}f}", reinterpret_cast<void*>(0xcafe), 2),
-      FormatError, "precision specifier requires floating-point argument");
-
-  EXPECT_THROW_MSG(format("{0:.{1}}", 'x', 2),
-      FormatError, "precision specifier requires floating-point argument");
-  EXPECT_THROW_MSG(format("{0:.{1}f}", 'x', 2),
-      FormatError, "precision specifier requires floating-point argument");
-
-  EXPECT_THROW_MSG(format("{0:.{1}}", "str", 2),
-      FormatError, "precision specifier requires floating-point argument");
-  EXPECT_THROW_MSG(format("{0:.{1}f}", "str", 2),
-      FormatError, "precision specifier requires floating-point argument");
-
-  EXPECT_THROW_MSG(format("{0:.{1}}", TestString(), 2),
-      FormatError, "precision specifier requires floating-point argument");
-  EXPECT_THROW_MSG(format("{0:.{1}f}", TestString(), 2),
-      FormatError, "precision specifier requires floating-point argument");
+      FormatError, "precision not allowed in pointer format specifier");
+
+  EXPECT_EQ("   ", format("{0:3.{1}}", 'x', 0));
+  EXPECT_EQ("st", format("{0:.{1}}", "str", 2));
+  EXPECT_EQ("te", format("{0:.{1}}", TestString("test"), 2));
 }
 
 template <typename T>