diff --git a/README.markdown b/README.markdown
index b76d16a..aeb9f4f 100644
--- a/README.markdown
+++ b/README.markdown
@@ -28,7 +28,8 @@ then simply copy and paste:
   the first file in a brand new Ruby project might very well be derived from
   your `.irbrc`.  I consider this a feature.
 * If your file is consistently indented with hard tabs, `'shiftwidth'` will be
-  set to your `'tabstop'`.  Otherwise, a `'tabstop'` of 8 is enforced.
+  set to 0 so that `'tabstop'` is used.  Otherwise, a `'tabstop'` of 8 is
+  enforced.
 * The algorithm is rolled from scratch, fairly simplistic, and only lightly
   battle tested.  It's probably not (yet) as good as [DetectIndent][].
   Let me know what it fails on for you.
diff --git a/plugin/sleuth.vim b/plugin/sleuth.vim
index 1904395..9a61da7 100644
--- a/plugin/sleuth.vim
+++ b/plugin/sleuth.vim
@@ -77,7 +77,11 @@ function! s:guess(lines) abort
   endfor
 
   if heuristics.hard && !heuristics.spaces
-    return {'expandtab': 0, 'shiftwidth': &tabstop}
+    if exists('*shiftwidth')
+      return {'expandtab': 0, 'shiftwidth': 0}
+    else
+      return {'expandtab': 0, 'shiftwidth': &tabstop}
+    endif
   elseif heuristics.soft != heuristics.hard
     let options.expandtab = heuristics.soft > heuristics.hard
     if heuristics.hard