diff --git a/tests/unit/browser/test_webelem.py b/tests/unit/browser/test_webelem.py
index 654bd3a7d..f36b16694 100644
--- a/tests/unit/browser/test_webelem.py
+++ b/tests/unit/browser/test_webelem.py
@@ -599,7 +599,7 @@ class TestJavascriptEscape:
     # http://qutebrowser.org:8010/builders/debian-jessie/builds/765/steps/unittests/
     # Should that be ignored?
 
-    @pytest.mark.parametrize('before, after', TESTS.items(), ids=repr)
+    @pytest.mark.parametrize('before, after', sorted(TESTS.items()), ids=repr)
     def test_fake_escape(self, before, after):
         """Test javascript escaping with some expected outcomes."""
         assert webelem.javascript_escape(before) == after
@@ -645,7 +645,7 @@ class TestJavascriptEscape:
         result = webframe.evaluateJavaScript('"{}";'.format(escaped))
         assert result == text
 
-    @pytest.mark.parametrize('text', TESTS, ids=repr)
+    @pytest.mark.parametrize('text', sorted(TESTS), ids=repr)
     def test_real_escape(self, webframe, qtbot, text):
         """Test javascript escaping with a real QWebPage."""
         self._test_escape(text, qtbot, webframe)
diff --git a/tests/unit/config/test_config.py b/tests/unit/config/test_config.py
index cb06364fd..a1d941137 100644
--- a/tests/unit/config/test_config.py
+++ b/tests/unit/config/test_config.py
@@ -189,7 +189,7 @@ class TestConfigParser:
         assert new_sect in configdata.DATA
 
     @pytest.mark.parametrize('old_tuple, new_option',
-        config.ConfigManager.RENAMED_OPTIONS.items())
+        sorted(config.ConfigManager.RENAMED_OPTIONS.items()))
     def test_renamed_options(self, old_tuple, new_option):
         """Make sure renamed options exist under the new name."""
         section, old_option = old_tuple
diff --git a/tests/unit/config/test_configtypes.py b/tests/unit/config/test_configtypes.py
index 34f4329fc..cacc898ee 100644
--- a/tests/unit/config/test_configtypes.py
+++ b/tests/unit/config/test_configtypes.py
@@ -254,7 +254,7 @@ class TestMappingType:
     def klass(self):
         return MappingSubclass
 
-    @pytest.mark.parametrize('val', TESTS.keys())
+    @pytest.mark.parametrize('val', sorted(TESTS.keys()))
     def test_validate_valid(self, klass, val):
         klass(none_ok=True).validate(val)
 
@@ -263,7 +263,7 @@ class TestMappingType:
         with pytest.raises(configexc.ValidationError):
             klass().validate(val)
 
-    @pytest.mark.parametrize('val, expected', TESTS.items())
+    @pytest.mark.parametrize('val, expected', sorted(TESTS.items()))
     def test_transform(self, klass, val, expected):
         assert klass().transform(val) == expected
 
@@ -488,11 +488,11 @@ class TestBool:
     def klass(self):
         return configtypes.Bool
 
-    @pytest.mark.parametrize('val, expected', TESTS.items())
+    @pytest.mark.parametrize('val, expected', sorted(TESTS.items()))
     def test_transform(self, klass, val, expected):
         assert klass().transform(val) == expected
 
-    @pytest.mark.parametrize('val', TESTS)
+    @pytest.mark.parametrize('val', sorted(TESTS))
     def test_validate_valid(self, klass, val):
         klass(none_ok=True).validate(val)
 
@@ -518,11 +518,11 @@ class TestBoolAsk:
     def klass(self):
         return configtypes.BoolAsk
 
-    @pytest.mark.parametrize('val, expected', TESTS.items())
+    @pytest.mark.parametrize('val, expected', sorted(TESTS.items()))
     def test_transform(self, klass, val, expected):
         assert klass().transform(val) == expected
 
-    @pytest.mark.parametrize('val', TESTS)
+    @pytest.mark.parametrize('val', sorted(TESTS))
     def test_validate_valid(self, klass, val):
         klass(none_ok=True).validate(val)
 
@@ -871,7 +871,7 @@ class TestColorSystem:
     def klass(self):
         return configtypes.ColorSystem
 
-    @pytest.mark.parametrize('val', TESTS)
+    @pytest.mark.parametrize('val', sorted(TESTS))
     def test_validate_valid(self, klass, val):
         klass(none_ok=True).validate(val)
 
@@ -880,7 +880,7 @@ class TestColorSystem:
         with pytest.raises(configexc.ValidationError):
             klass().validate(val)
 
-    @pytest.mark.parametrize('val, expected', TESTS.items())
+    @pytest.mark.parametrize('val, expected', sorted(TESTS.items()))
     def test_transform(self, klass, val, expected):
         assert klass().transform(val) == expected
 
@@ -1048,7 +1048,7 @@ class TestFont:
     def qtfont_class(self):
         return configtypes.QtFont
 
-    @pytest.mark.parametrize('val', list(TESTS) + [''])
+    @pytest.mark.parametrize('val', sorted(list(TESTS)) + [''])
     def test_validate_valid(self, klass, val):
         klass(none_ok=True).validate(val)
 
@@ -1077,11 +1077,11 @@ class TestFont:
         with pytest.raises(configexc.ValidationError):
             klass().validate('')
 
-    @pytest.mark.parametrize('string', TESTS)
+    @pytest.mark.parametrize('string', sorted(TESTS))
     def test_transform_font(self, font_class, string):
         assert font_class().transform(string) == string
 
-    @pytest.mark.parametrize('string, desc', TESTS.items())
+    @pytest.mark.parametrize('string, desc', sorted(TESTS.items()))
     def test_transform_qtfont(self, qtfont_class, string, desc):
         assert Font(qtfont_class().transform(string)) == Font.fromdesc(desc)
 
@@ -1829,7 +1829,7 @@ class TestAutoSearch:
     def klass(self):
         return configtypes.AutoSearch
 
-    @pytest.mark.parametrize('val', TESTS)
+    @pytest.mark.parametrize('val', sorted(TESTS))
     def test_validate_valid(self, klass, val):
         klass(none_ok=True).validate(val)
 
@@ -1838,7 +1838,7 @@ class TestAutoSearch:
         with pytest.raises(configexc.ValidationError):
             klass().validate(val)
 
-    @pytest.mark.parametrize('val, expected', TESTS.items())
+    @pytest.mark.parametrize('val, expected', sorted(TESTS.items()))
     def test_transform(self, klass, val, expected):
         assert klass().transform(val) == expected
 
@@ -1859,7 +1859,7 @@ class TestIgnoreCase:
     def klass(self):
         return configtypes.IgnoreCase
 
-    @pytest.mark.parametrize('val', TESTS)
+    @pytest.mark.parametrize('val', sorted(TESTS))
     def test_validate_valid(self, klass, val):
         klass(none_ok=True).validate(val)
 
@@ -1868,7 +1868,7 @@ class TestIgnoreCase:
         with pytest.raises(configexc.ValidationError):
             klass().validate(val)
 
-    @pytest.mark.parametrize('val, expected', TESTS.items())
+    @pytest.mark.parametrize('val, expected', sorted(TESTS.items()))
     def test_transform(self, klass, val, expected):
         assert klass().transform(val) == expected
 
@@ -1910,7 +1910,7 @@ class TestUrlList:
     def klass(self):
         return configtypes.UrlList
 
-    @pytest.mark.parametrize('val', TESTS)
+    @pytest.mark.parametrize('val', sorted(TESTS))
     def test_validate_valid(self, klass, val):
         klass(none_ok=True).validate(val)
 
@@ -1928,7 +1928,7 @@ class TestUrlList:
         with pytest.raises(configexc.ValidationError):
             klass().validate('foo,,bar')
 
-    @pytest.mark.parametrize('val, expected', TESTS.items())
+    @pytest.mark.parametrize('val, expected', sorted(TESTS.items()))
     def test_transform_single(self, klass, val, expected):
         assert klass().transform(val) == expected
 
@@ -1968,7 +1968,7 @@ class TestConfirmQuit:
     def klass(self):
         return configtypes.ConfirmQuit
 
-    @pytest.mark.parametrize('val', TESTS.keys())
+    @pytest.mark.parametrize('val', sorted(TESTS.keys()))
     def test_validate_valid(self, klass, val):
         klass(none_ok=True).validate(val)
 
@@ -1985,7 +1985,7 @@ class TestConfirmQuit:
         with pytest.raises(configexc.ValidationError):
             klass().validate(val)
 
-    @pytest.mark.parametrize('val, expected', TESTS.items())
+    @pytest.mark.parametrize('val, expected', sorted(TESTS.items()))
     def test_transform(self, klass, val, expected):
         assert klass().transform(val) == expected
 
diff --git a/tests/unit/misc/test_crashdialog.py b/tests/unit/misc/test_crashdialog.py
index e44bf3559..432bb721e 100644
--- a/tests/unit/misc/test_crashdialog.py
+++ b/tests/unit/misc/test_crashdialog.py
@@ -94,7 +94,7 @@ class TestParseFatalStacktrace:
         "QT_IM_MODULE = fcitx"
     ),
     ({'LANGUAGE': 'foo', 'LANG': 'en_US.UTF-8'}, "LANG = en_US.UTF-8"),
-], ids=repr)
+], ids=lambda e: e[1])
 def test_get_environment_vars(monkeypatch, env, expected):
     """Test for crashdialog._get_environment_vars."""
     for key in os.environ.copy():
diff --git a/tests/unit/misc/test_split.py b/tests/unit/misc/test_split.py
index 265d62079..272b281ea 100644
--- a/tests/unit/misc/test_split.py
+++ b/tests/unit/misc/test_split.py
@@ -126,7 +126,8 @@ class TestSplit:
 
     """Test split."""
 
-    @pytest.fixture(params=_parse_split_test_data_str(), ids=lambda e: e.input)
+    @pytest.fixture(params=list(_parse_split_test_data_str()),
+                    ids=lambda e: e.input)
     def split_test_case(self, request):
         """Fixture to automatically parametrize all depending tests.
 
@@ -163,7 +164,7 @@ class TestSimpleSplit:
         'foo\nbar': ['foo', '\nbar'],
     }
 
-    @pytest.mark.parametrize('test', TESTS, ids=repr)
+    @pytest.mark.parametrize('test', sorted(TESTS), ids=repr)
     def test_str_split(self, test):
         """Test if the behavior matches str.split."""
         assert split.simple_split(test) == test.rstrip().split()
@@ -177,7 +178,7 @@ class TestSimpleSplit:
         expected = s.rstrip().split(maxsplit=maxsplit)
         assert actual == expected
 
-    @pytest.mark.parametrize('test, expected', TESTS.items(), ids=repr)
+    @pytest.mark.parametrize('test, expected', sorted(TESTS.items()), ids=repr)
     def test_split_keep(self, test, expected):
         """Test splitting with keep=True."""
         assert split.simple_split(test, keep=True) == expected
diff --git a/tests/unit/utils/overflow_test_cases.py b/tests/unit/utils/overflow_test_cases.py
index e209e1827..ff1d73f03 100644
--- a/tests/unit/utils/overflow_test_cases.py
+++ b/tests/unit/utils/overflow_test_cases.py
@@ -49,12 +49,20 @@ BAD_VALUES = {
 }
 
 
+def good_values():
+    return list(iter_good_values())
+
+
+def bad_values():
+    return list(iter_bad_values())
+
+
 def iter_good_values():
     """Yield "good" (C data type, value) tuples.
 
     Those should pass overflow checking.
     """
-    for ctype, values in GOOD_VALUES.items():
+    for ctype, values in sorted(GOOD_VALUES.items()):
         for value in values:
             yield ctype, value
 
@@ -65,6 +73,6 @@ def iter_bad_values():
     These should not pass overflow checking. The third value is the value they
     should be replaced with if overflow checking should not be fatal.
     """
-    for ctype, values in BAD_VALUES.items():
+    for ctype, values in sorted(BAD_VALUES.items()):
         for value, repl in values:
             yield ctype, value, repl
diff --git a/tests/unit/utils/test_qtutils.py b/tests/unit/utils/test_qtutils.py
index 18c15d04a..a429784d0 100644
--- a/tests/unit/utils/test_qtutils.py
+++ b/tests/unit/utils/test_qtutils.py
@@ -69,21 +69,21 @@ class TestCheckOverflow:
     """Test check_overflow."""
 
     @pytest.mark.parametrize('ctype, val',
-                             overflow_test_cases.iter_good_values())
+                             overflow_test_cases.good_values())
     def test_good_values(self, ctype, val):
         """Test values which are inside bounds."""
         qtutils.check_overflow(val, ctype)
 
     @pytest.mark.parametrize('ctype, val',
                              [(ctype, val) for (ctype, val, _) in
-                              overflow_test_cases.iter_bad_values()])
+                              overflow_test_cases.bad_values()])
     def test_bad_values_fatal(self, ctype, val):
         """Test values which are outside bounds with fatal=True."""
         with pytest.raises(OverflowError):
             qtutils.check_overflow(val, ctype)
 
     @pytest.mark.parametrize('ctype, val, repl',
-                             overflow_test_cases.iter_bad_values())
+                             overflow_test_cases.bad_values())
     def test_bad_values_nonfatal(self, ctype, val, repl):
         """Test values which are outside bounds with fatal=False."""
         newval = qtutils.check_overflow(val, ctype, fatal=False)
diff --git a/tests/unit/utils/test_utils.py b/tests/unit/utils/test_utils.py
index da40a43d7..26b3c0920 100644
--- a/tests/unit/utils/test_utils.py
+++ b/tests/unit/utils/test_utils.py
@@ -805,7 +805,8 @@ QUALNAME_OBJ = QualnameObj()
     (qutebrowser, 'qutebrowser'),  # module
     (qutebrowser.utils, 'qutebrowser.utils'),  # submodule
     (utils, 'qutebrowser.utils.utils'),  # submodule (from-import)
-])
+], ids=['instance', 'class', 'unbound-method', 'bound-method', 'function',
+        'partial', 'module', 'submodule', 'from-import'])
 def test_qualname(obj, expected):
     assert utils.qualname(obj) == expected