diff --git a/nikola/plugins/task/sitemap/__init__.py b/nikola/plugins/task/sitemap/__init__.py
index 641b41d1c9..e7515a39c2 100644
--- a/nikola/plugins/task/sitemap/__init__.py
+++ b/nikola/plugins/task/sitemap/__init__.py
@@ -40,7 +40,7 @@
     import urllib.robotparser as robotparser  # NOQA
 
 from nikola.plugin_categories import LateTask
-from nikola.utils import apply_filters, config_changed, encodelink
+from nikola.utils import apply_filters, config_changed, encodelink, get_asset_path
 
 
 urlset_header = """<?xml version="1.0" encoding="UTF-8"?>
@@ -118,6 +118,7 @@ def gen_tasks(self):
             "base_url": self.site.config["BASE_URL"],
             "site_url": self.site.config["SITE_URL"],
             "output_folder": self.site.config["OUTPUT_FOLDER"],
+            "files_folders": self.site.config['FILES_FOLDERS'],
             "strip_indexes": self.site.config["STRIP_INDEXES"],
             "index_file": self.site.config["INDEX_FILE"],
             "sitemap_include_fileless_dirs": self.site.config["SITEMAP_INCLUDE_FILELESS_DIRS"],
@@ -140,9 +141,8 @@ def gen_tasks(self):
         sitemapindex = {}
         urlset = {}
 
-        def scan_locs():
+        def scan_locs(robots_rules):
             """Scan site locations."""
-            robots_rules = parse_robots_exclusions(kw['robots_exclusions'])
             for root, dirs, files in os.walk(output, followlinks=True):
                 if not dirs and not files and not kw['sitemap_include_fileless_dirs']:
                     continue  # Totally empty, not on sitemap
@@ -228,6 +228,16 @@ def scan_locs():
                                 alternates.append(alternates_format.format(lang, alt_url))
                         urlset[loc] = loc_format.format(encodelink(loc), lastmod, '\n'.join(alternates))
 
+        def parse_robotstxt(path):
+            robot = robotparser.RobotFileParser()
+            fh = io.open(path, 'r', encoding='utf-8-sig')
+            rules = fh.readlines()
+            if sys.version_info[0] == 2:
+                rules = [ line.encode('utf-8') for line in rules ]
+            fh.close()
+            robot.parse(rules)
+            return robot
+
         def parse_robots_exclusions(exclusions):
             """Parse rules to check fetchable."""
             rules = []
@@ -268,7 +278,12 @@ def scan_locs_task():
             Other tasks can depend on this output, instead of having
             to scan locations.
             """
-            scan_locs()
+            robotstxt = get_asset_path("robots.txt", [], files_folders=kw["files_folders"])
+            if robotstxt:
+                robots_rules = parse_robotstxt(robotstxt)
+            else:
+                robots_rules = parse_robots_exclusions(kw['robots_exclusions'])
+            scan_locs(robots_rules)
 
             # Generate a list of file dependencies for the actual generation
             # task, so rebuilds are triggered.  (Issue #1032)
@@ -289,6 +304,9 @@ def scan_locs_task():
                 if os.path.isdir(p) and os.path.exists(os.path.join(p, 'index.html')):
                     file_dep.append(p + 'index.html')
 
+            if robotstxt:
+                file_dep.append(os.path.join(output, 'robots.txt'))
+
             return {'file_dep': file_dep}
 
         yield {