diff --git a/src/doc/book/strings.md b/src/doc/book/strings.md
index 6af15d8768363..a2146b669e3c6 100644
--- a/src/doc/book/strings.md
+++ b/src/doc/book/strings.md
@@ -163,8 +163,8 @@ let hachi = &dog[0..2];
 with this error:
 
 ```text
-thread 'main' panicked at 'index 0 and/or 2 in `忠犬ハチ公` do not lie on
-character boundary'
+thread 'main' panicked at 'byte index 2 is not a char boundary; it is inside '忠'
+(bytes 0..3) of `忠犬ハチ公`'
 ```
 
 ## Concatenation
diff --git a/src/libcollectionstest/str.rs b/src/libcollectionstest/str.rs
index 14a0819d381bc..9c3c3740aa96f 100644
--- a/src/libcollectionstest/str.rs
+++ b/src/libcollectionstest/str.rs
@@ -383,17 +383,29 @@ tempus vel, gravida nec quam.";
 
 // check the panic includes the prefix of the sliced string
 #[test]
-#[should_panic(expected="Lorem ipsum dolor sit amet")]
+#[should_panic(expected="byte index 1024 is out of bounds of `Lorem ipsum dolor sit amet")]
 fn test_slice_fail_truncated_1() {
     &LOREM_PARAGRAPH[..1024];
 }
 // check the truncation in the panic message
 #[test]
-#[should_panic(expected="luctus, im`[...] do not lie on character boundary")]
+#[should_panic(expected="luctus, im`[...]")]
 fn test_slice_fail_truncated_2() {
     &LOREM_PARAGRAPH[..1024];
 }
 
+#[test]
+#[should_panic(expected="byte index 4 is not a char boundary; it is inside 'α' (bytes 3..5) of")]
+fn test_slice_fail_boundary_1() {
+    &"abcαβγ"[4..];
+}
+
+#[test]
+#[should_panic(expected="byte index 6 is not a char boundary; it is inside 'β' (bytes 5..7) of")]
+fn test_slice_fail_boundary_2() {
+    &"abcαβγ"[2..6];
+}
+
 #[test]
 fn test_slice_from() {
     assert_eq!(&"abcd"[0..], "abcd");
diff --git a/src/libcore/str/mod.rs b/src/libcore/str/mod.rs
index b4cd52e59f658..7081c3ebf43a0 100644
--- a/src/libcore/str/mod.rs
+++ b/src/libcore/str/mod.rs
@@ -1741,13 +1741,31 @@ fn truncate_to_char_boundary(s: &str, mut max: usize) -> (bool, &str) {
 #[cold]
 fn slice_error_fail(s: &str, begin: usize, end: usize) -> ! {
     const MAX_DISPLAY_LENGTH: usize = 256;
-    let (truncated, s) = truncate_to_char_boundary(s, MAX_DISPLAY_LENGTH);
+    let (truncated, s_trunc) = truncate_to_char_boundary(s, MAX_DISPLAY_LENGTH);
     let ellipsis = if truncated { "[...]" } else { "" };
 
+    // 1. out of bounds
+    if begin > s.len() || end > s.len() {
+        let oob_index = if begin > s.len() { begin } else { end };
+        panic!("byte index {} is out of bounds of `{}`{}", oob_index, s_trunc, ellipsis);
+    }
+
+    // 2. begin <= end
     assert!(begin <= end, "begin <= end ({} <= {}) when slicing `{}`{}",
-            begin, end, s, ellipsis);
-    panic!("index {} and/or {} in `{}`{} do not lie on character boundary",
-          begin, end, s, ellipsis);
+            begin, end, s_trunc, ellipsis);
+
+    // 3. character boundary
+    let index = if !s.is_char_boundary(begin) { begin } else { end };
+    // find the character
+    let mut char_start = index;
+    while !s.is_char_boundary(char_start) {
+        char_start -= 1;
+    }
+    // `char_start` must be less than len and a char boundary
+    let ch = s[char_start..].chars().next().unwrap();
+    let char_range = char_start .. char_start + ch.len_utf8();
+    panic!("byte index {} is not a char boundary; it is inside {:?} (bytes {:?}) of `{}`{}",
+           index, ch, char_range, s_trunc, ellipsis);
 }
 
 #[stable(feature = "core", since = "1.6.0")]