diff --git a/src/runtime/object.cpp b/src/runtime/object.cpp
index e95f9efaa1..a9c92ec7da 100644
--- a/src/runtime/object.cpp
+++ b/src/runtime/object.cpp
@@ -1593,11 +1593,23 @@ usize string_utf8_next(b_obj_arg s, usize i) {
     return i + 1;
 }
 
+static inline bool is_utf8_first_byte(unsigned char c) {
+    return (c & 0x80) == 0 || (c & 0xe0) == 0xc0 || (c & 0xf0) == 0xe0 || (c & 0xf8) == 0xf0;
+}
+
 obj_res string_utf8_extract(b_obj_arg s, usize b, usize e) {
+    char const * str = string_cstr(s);
     usize sz = string_size(s) - 1;
     if (b >= e || b >= sz) return mk_string("");
+    /* In the reference implementation if `b` is not pointing to a valid UTF8
+       character start position, the result is the empty string. */
+    if (!is_utf8_first_byte(str[b])) return mk_string("");
     if (e > sz) e = sz;
     lean_assert(b < e);
+    lean_assert(e > 0);
+    /* In the reference implementation if `e` is not pointing to a valid UTF8
+       character start position, it is assumed to be at the end. */
+    if (e < sz && !is_utf8_first_byte(str[e])) e = sz;
     usize new_sz = e - b;
     lean_assert(new_sz > 0);
     obj_res r = alloc_string(new_sz+1, new_sz+1, 0);
diff --git a/src/runtime/utf8.cpp b/src/runtime/utf8.cpp
index c8cff31369..4729f28f31 100644
--- a/src/runtime/utf8.cpp
+++ b/src/runtime/utf8.cpp
@@ -29,7 +29,7 @@ unsigned get_utf8_size(unsigned char c) {
     else if (c == 0xFF)
         return 1;
     else
-        return 0;
+        return 1; /* invalid */
 }
 
 size_t utf8_strlen(char const * str) {
diff --git a/tests/compiler/str.lean b/tests/compiler/str.lean
index c6bbdef34f..26c6efae67 100644
--- a/tests/compiler/str.lean
+++ b/tests/compiler/str.lean
@@ -6,7 +6,15 @@ def show_chars : nat → string → string.utf8_pos → io unit
     show_chars n s (s.utf8_next idx)
 
 def main : io uint32 :=
-let s₁ := "hello α world β" in
-io.println' (to_string s₁.utf8_byte_size) *>
+let s₁ := "hello α_world_β" in
+let b  := string.utf8_begin in
+let e  := s₁.utf8_byte_size in
+io.println' (s₁.utf8_extract b e) *>
+io.println' (s₁.utf8_extract (b+2) e) *>
+io.println' (s₁.utf8_extract (b+2) (e-1)) *>
+io.println' (s₁.utf8_extract (b+2) (e-2)) *>
+io.println' (s₁.utf8_extract (b+7) e) *>
+io.println' (s₁.utf8_extract (b+8) e) *>
+io.println' (to_string e) *>
 show_chars s₁.utf8_byte_size.to_nat s₁ 0  *>
 pure 0
diff --git a/tests/compiler/str.lean.expected.out b/tests/compiler/str.lean.expected.out
index 9059a5ca03..f02fe3977e 100644
--- a/tests/compiler/str.lean.expected.out
+++ b/tests/compiler/str.lean.expected.out
@@ -1,3 +1,9 @@
+hello α_world_β
+llo α_world_β
+llo α_world_β
+llo α_world_
+
+_world_β
 17
 >> h
 >> e
@@ -6,11 +12,11 @@
 >> o
 >>  
 >> α
->>  
+>> _
 >> w
 >> o
 >> r
 >> l
 >> d
->>  
+>> _
 >> β