diff --git a/sh_scrapy/extension.py b/sh_scrapy/extension.py
index 00dc758..2f027da 100644
--- a/sh_scrapy/extension.py
+++ b/sh_scrapy/extension.py
@@ -1,5 +1,6 @@
 import logging
 from contextlib import suppress
+from warnings import warn
 from weakref import WeakKeyDictionary
 
 import scrapy
@@ -9,7 +10,6 @@
 from scrapy.exporters import PythonItemExporter
 from scrapy.http import Request
 from scrapy.utils.deprecate import create_deprecated_class
-from scrapy.utils.request import request_fingerprint
 
 from sh_scrapy import hsref
 from sh_scrapy.crawl import ignore_warnings
@@ -79,13 +79,41 @@ def spider_closed(self, spider, reason):
 """
 
 
-class HubstorageMiddleware(object):
+class HubstorageMiddleware:
 
-    def __init__(self):
+    @classmethod
+    def from_crawler(cls, crawler):
+        try:
+            result = cls(crawler)
+        except TypeError:
+            warn(
+                (
+                    "Subclasses of HubstorageMiddleware must now accept a "
+                    "crawler parameter in their __init__ method. This will "
+                    "become an error in the future."
+                ),
+                DeprecationWarning,
+            )
+            result = cls()
+            result._crawler = crawler
+            result._load_fingerprinter()
+        return result
+
+    def _load_fingerprinter(self):
+        if hasattr(self._crawler, "request_fingerprinter"):
+            self._fingerprint = lambda request: self._crawler.request_fingerprinter.fingerprint(request).hex()
+        else:
+            from scrapy.utils.request import request_fingerprint
+            self._fingerprint = request_fingerprint
+
+    def __init__(self, crawler=None):
         self._seen = WeakKeyDictionary()
         self.hsref = hsref.hsref
         self.pipe_writer = pipe_writer
         self.request_id_sequence = request_id_sequence
+        self._crawler = crawler
+        if crawler:
+            self._load_fingerprinter()
 
     def process_spider_input(self, response, spider):
         self.pipe_writer.write_request(
@@ -95,7 +123,7 @@ def process_spider_input(self, response, spider):
             rs=len(response.body),
             duration=response.meta.get('download_latency', 0) * 1000,
             parent=response.meta.get(HS_PARENT_ID_KEY),
-            fp=request_fingerprint(response.request),
+            fp=self._fingerprint(response.request),
         )
         self._seen[response] = next(self.request_id_sequence)
 
diff --git a/tests/test_extension.py b/tests/test_extension.py
index 067e3cc..71fab7d 100644
--- a/tests/test_extension.py
+++ b/tests/test_extension.py
@@ -3,11 +3,13 @@
 
 import mock
 import pytest
+import scrapy
+from packaging import version
+from pytest import warns
 from scrapy import Spider
 from scrapy.exporters import PythonItemExporter
 from scrapy.http import Request, Response
 from scrapy.item import Item
-from scrapy.utils.request import request_fingerprint
 from scrapy.utils.test import get_crawler
 
 from sh_scrapy.extension import HubstorageExtension, HubstorageMiddleware
@@ -92,7 +94,8 @@ def test_hs_ext_spider_closed(hs_ext):
 @pytest.fixture
 def hs_mware(monkeypatch):
     monkeypatch.setattr('sh_scrapy.extension.pipe_writer', mock.Mock())
-    return HubstorageMiddleware()
+    crawler = get_crawler()
+    return HubstorageMiddleware.from_crawler(crawler)
 
 
 def test_hs_mware_init(hs_mware):
@@ -106,9 +109,13 @@ def test_hs_mware_process_spider_input(hs_mware):
     hs_mware.process_spider_input(response, Spider('test'))
     assert hs_mware.pipe_writer.write_request.call_count == 1
     args = hs_mware.pipe_writer.write_request.call_args[1]
+    if hasattr(hs_mware._crawler, "request_fingerprinter"):
+        fp = "1c735665b072000e11b0169081bce5bbaeac09a7"
+    else:
+        fp = "a001a1eb4537acdc8525edf1250065cab2657152"
     assert args == {
         'duration': 0,
-        'fp': request_fingerprint(response.request),
+        'fp': fp,
         'method': 'GET',
         'parent': None,
         'rs': 0,
@@ -138,3 +145,40 @@ def test_hs_mware_process_spider_output_filter_request(hs_mware):
     # make sure that we update hsparent meta only for requests
     assert result[0].meta.get(HS_PARENT_ID_KEY) is None
     assert result[1].meta[HS_PARENT_ID_KEY] == 'riq'
+
+
+@pytest.mark.skipif(
+    version.parse(scrapy.__version__) < version.parse("2.7"),
+    reason="Only Scrapy 2.7 and higher support centralized request fingerprints."
+)
+def test_custom_fingerprinter(monkeypatch):
+    monkeypatch.setattr('sh_scrapy.extension.pipe_writer', mock.Mock())
+
+    class CustomFingerprinter:
+        def fingerprint(self, request):
+            return b"foo"
+
+    crawler = get_crawler(settings_dict={"REQUEST_FINGERPRINTER_CLASS": CustomFingerprinter})
+    mw = HubstorageMiddleware.from_crawler(crawler)
+
+    response = Response('http://resp-url')
+    response.request = Request('http://req-url')
+    mw.process_spider_input(response, Spider('test'))
+    assert mw.pipe_writer.write_request.call_args[1]["fp"] == b"foo".hex()
+
+
+def test_subclassing():
+    class CustomHubstorageMiddleware(HubstorageMiddleware):
+        def __init__(self):
+            super().__init__()
+            self.foo = "bar"
+
+    crawler = get_crawler()
+    with warns(
+        DeprecationWarning,
+        match="must now accept a crawler parameter in their __init__ method",
+    ):
+        mw = CustomHubstorageMiddleware.from_crawler(crawler)
+
+    assert mw.foo == "bar"
+    assert hasattr(mw, "_fingerprint")
diff --git a/tox.ini b/tox.ini
index e6b5829..add9095 100644
--- a/tox.ini
+++ b/tox.ini
@@ -1,6 +1,9 @@
 # tox.ini
 [tox]
 envlist = py36-scrapy16, py
+requires =
+    # https://github.com/pypa/virtualenv/issues/2550
+    virtualenv<=20.21.1
 
 [testenv]
 deps =