diff --git a/interfaces/cython/cantera/test/test_composite.py b/interfaces/cython/cantera/test/test_composite.py
index 2667bec4459..ac86613edcc 100644
--- a/interfaces/cython/cantera/test/test_composite.py
+++ b/interfaces/cython/cantera/test/test_composite.py
@@ -1,5 +1,4 @@
 import sys
-from pathlib import Path
 
 import numpy as np
 from collections import OrderedDict
@@ -15,7 +14,7 @@ class TestModels(utilities.CanteraTest):
     @classmethod
     def setUpClass(cls):
         utilities.CanteraTest.setUpClass()
-        cls.yml_file = Path(cls.test_data_dir).joinpath("thermo-models.yaml")
+        cls.yml_file = cls.test_data_path / "thermo-models.yaml"
         cls.yml = utilities.load_yaml(cls.yml_file)
 
     def test_load_thermo_models(self):
@@ -125,13 +124,6 @@ def setUpClass(cls):
         utilities.CanteraTest.setUpClass()
         cls.gas = ct.Solution('h2o2.yaml', transport_model=None)
 
-    def tearDown(self):
-        temp_files = ["solutionarray.csv", "solutionarray.h5"]
-        for f in temp_files:
-            fpath = Path(self.test_work_dir).joinpath(f)
-            if fpath.is_file():
-                fpath.unlink()
-
     def test_collect_data(self):
         states = ct.SolutionArray(self.gas)
         collected = states.collect_data(tabular=True)
@@ -149,7 +141,7 @@ def test_write_csv(self):
         states.TPX = np.linspace(300, 1000, 7), 2e5, 'H2:0.5, O2:0.4'
         states.equilibrate('HP')
 
-        outfile = Path(self.test_work_dir).joinpath("solutionarray.csv")
+        outfile = self.test_work_path / "solutionarray.csv"
         states.write_csv(outfile)
 
         data = np.genfromtxt(outfile, names=True, delimiter=',')
@@ -166,7 +158,7 @@ def test_write_csv(self):
     def test_write_csv_str_column(self):
         states = ct.SolutionArray(self.gas, 3, extra={'spam': 'eggs'})
 
-        outfile = Path(self.test_work_dir).joinpath("solutionarray.csv")
+        outfile = self.test_work_path / "solutionarray.csv"
         states.write_csv(outfile)
 
         b = ct.SolutionArray(self.gas, extra={'spam'})
@@ -176,7 +168,7 @@ def test_write_csv_str_column(self):
     def test_write_csv_multidim_column(self):
         states = ct.SolutionArray(self.gas, 3, extra={'spam': np.zeros((3, 5,))})
 
-        outfile = Path(self.test_work_dir).joinpath("solutionarray.csv")
+        outfile = self.test_work_path / "solutionarray.csv"
         with self.assertRaisesRegex(NotImplementedError, 'not supported'):
             states.write_csv(outfile)
 
@@ -193,7 +185,9 @@ def test_to_pandas(self):
 
     @utilities.unittest.skipIf(isinstance(_h5py, ImportError), "h5py is not installed")
     def test_write_hdf(self):
-        outfile = Path(self.test_work_dir).joinpath("solutionarray.h5")
+        outfile = self.test_work_path / "solutionarray.h5"
+        if outfile.is_file():
+            outfile.unlink()
 
         extra = {'foo': range(7), 'bar': range(7)}
         meta = {'spam': 'eggs', 'hello': 'world'}
@@ -236,7 +230,9 @@ def test_write_hdf(self):
 
     @utilities.unittest.skipIf(isinstance(_h5py, ImportError), "h5py is not installed")
     def test_write_hdf_str_column(self):
-        outfile = Path(self.test_work_dir).joinpath("solutionarray.h5")
+        outfile = self.test_work_path / "solutionarray.h5"
+        if outfile.is_file():
+            outfile.unlink()
 
         states = ct.SolutionArray(self.gas, 3, extra={'spam': 'eggs'})
         states.write_hdf(outfile, mode='w')
@@ -247,7 +243,9 @@ def test_write_hdf_str_column(self):
 
     @utilities.unittest.skipIf(isinstance(_h5py, ImportError), "h5py is not installed")
     def test_write_hdf_multidim_column(self):
-        outfile = Path(self.test_work_dir).joinpath("solutionarray.h5")
+        outfile = self.test_work_path / "solutionarray.h5"
+        if outfile.is_file():
+            outfile.unlink()
 
         states = ct.SolutionArray(self.gas, 3, extra={'spam': [[1, 2], [3, 4], [5, 6]]})
         states.write_hdf(outfile, mode='w')
@@ -496,7 +494,7 @@ def test_yaml_outunits(self):
         units = {'length': 'cm', 'quantity': 'mol', 'energy': 'cal'}
         gas.write_yaml('h2o2-generated.yaml', units=units)
         generated = utilities.load_yaml("h2o2-generated.yaml")
-        original = utilities.load_yaml(Path(self.cantera_data).joinpath("h2o2.yaml"))
+        original = utilities.load_yaml(self.cantera_data_path / "h2o2.yaml")
         self.assertEqual(generated['units'], units)
 
         for r1, r2 in zip(original['reactions'], generated['reactions']):
diff --git a/interfaces/cython/cantera/test/test_convert.py b/interfaces/cython/cantera/test/test_convert.py
index 4a71531e4d2..1aed50618c5 100644
--- a/interfaces/cython/cantera/test/test_convert.py
+++ b/interfaces/cython/cantera/test/test_convert.py
@@ -15,16 +15,16 @@ def convert(self, inputFile, thermo=None, transport=None,
         if output is None:
             output = Path(inputFile).stem  # strip '.inp'
         if inputFile is not None:
-            inputFile = Path(self.test_data_dir).joinpath(inputFile)
+            inputFile = self.test_data_path / inputFile
         if thermo is not None:
-            thermo = Path(self.test_data_dir).joinpath(thermo)
+            thermo = self.test_data_path / thermo
         if transport is not None:
-            transport = Path(self.test_data_dir).joinpath(transport)
+            transport = self.test_data_path / transport
         if surface is not None:
-            surface = Path(self.test_data_dir).joinpath(surface)
+            surface = self.test_data_path / surface
         if extra is not None:
-            extra = Path(self.test_data_dir).joinpath(extra)
-        output = Path(self.test_work_dir).joinpath(output + self.ext)
+            extra = self.test_data_path / extra
+        output = self.test_work_path / (output + self.ext)
         if output.is_file():
             output.unlink()
         self._convert(inputFile, thermo=thermo, transport=transport,
@@ -359,14 +359,14 @@ def test_empty_reaction_section(self):
 
     def test_reaction_comments1(self):
         output = self.convert('pdep-test.inp')
-        text = Path(output).read_text()
+        text = output.read_text()
         self.assertIn('Generic mechanism header', text)
         self.assertIn('Single PLOG reaction', text)
         self.assertIn('Multiple PLOG expressions at the same pressure', text)
 
     def test_reaction_comments2(self):
         output = self.convert('explicit-third-bodies.inp', thermo='dummy-thermo.dat')
-        text = Path(output).read_text()
+        text = output.read_text()
         self.assertIn('An end of line comment', text)
         self.assertIn('A comment after the last reaction', text)
 
@@ -470,7 +470,7 @@ def test_extra(self):
                      transport='gri30_tran.dat', output='gri30_extra',
                      extra='extra.yaml')
 
-        output = Path(self.test_work_dir).joinpath("gri30_extra" + self.ext)
+        output = self.test_work_path / ("gri30_extra" + self.ext)
         yml = utilities.load_yaml(output)
 
         desc = yml['description'].split('\n')[-1]
@@ -481,7 +481,7 @@ def test_extra(self):
 
     def test_sri_zero(self):
         self.convert('sri_convert_test.txt')
-        output = Path(self.test_work_dir).joinpath("sri_convert_test" + self.ext)
+        output = self.test_work_path / ("sri_convert_test" + self.ext)
         mech = utilities.load_yaml(output)
         D = mech['reactions'][0]['SRI']['D']
         E = mech['reactions'][0]['SRI']['E']
@@ -569,7 +569,7 @@ def test_long_source_input(self):
 
         gas = ct.Solution('pdep-test.yaml')
 
-        with open(Path(self.test_data_dir).joinpath("pdep-test.cti"), "r") as f:
+        with open(self.test_data_path / "pdep-test.cti", "r") as f:
             data = f.read()
         data_size_2048kB = data + ' '*2048*1024
         gas2 = ct.Solution(source=data_size_2048kB)
@@ -584,7 +584,7 @@ def test_short_source_input(self):
 
         gas = ct.Solution('pdep-test.yaml')
 
-        with open(Path(self.test_data_dir).joinpath("pdep-test.cti"), "r") as f:
+        with open(self.test_data_path / "pdep-test.cti", "r") as f:
             data = f.read()
         data_size_32kB = data + ' '*18000
         gas2 = ct.Solution(source=data_size_32kB)
diff --git a/interfaces/cython/cantera/test/test_equilibrium.py b/interfaces/cython/cantera/test/test_equilibrium.py
index eea5cbb76ff..f0a236e580b 100644
--- a/interfaces/cython/cantera/test/test_equilibrium.py
+++ b/interfaces/cython/cantera/test/test_equilibrium.py
@@ -1,5 +1,4 @@
 import unittest
-from os.path import join as pjoin
 
 import numpy as np
 
@@ -153,7 +152,7 @@ def test_equil_TP(self):
 
             data[i,1:] = self.mix.species_moles
 
-        self.compare(data, pjoin(self.test_data_dir, 'koh-equil-TP.csv'))
+        self.compare(data, self.test_data_path / "koh-equil-TP.csv")
 
     @utilities.slow_test
     def test_equil_HP(self):
@@ -177,7 +176,7 @@ def test_equil_HP(self):
             data[i,1] = self.mix.T # equilibrated temperature
             data[i,2:] = self.mix.species_moles
 
-        self.compare(data, pjoin(self.test_data_dir, 'koh-equil-HP.csv'))
+        self.compare(data, self.test_data_path / "koh-equil-HP.csv")
 
 
 class TestEquil_GasCarbon(utilities.CanteraTest):
@@ -207,7 +206,7 @@ def solve(self, solver, **kwargs):
             data[i,:2] = (phi[i], mix.T)
             data[i,2:] = mix.species_moles
 
-        self.compare(data, pjoin(self.test_data_dir, 'gas-carbon-equil.csv'))
+        self.compare(data, self.test_data_path / "gas-carbon-equil.csv")
 
     @utilities.slow_test
     def test_gibbs(self):
diff --git a/interfaces/cython/cantera/test/test_kinetics.py b/interfaces/cython/cantera/test/test_kinetics.py
index 55000f3c039..96f08863ee9 100644
--- a/interfaces/cython/cantera/test/test_kinetics.py
+++ b/interfaces/cython/cantera/test/test_kinetics.py
@@ -1,8 +1,6 @@
 import numpy as np
 import re
 import itertools
-from os.path import join as pjoin
-import os
 
 import cantera as ct
 from . import utilities
@@ -816,7 +814,7 @@ def cathode_curr(E):
                              cathode_bulk.electric_potential -
                              anode_bulk.electric_potential])
 
-        self.compare(data, pjoin(self.test_data_dir, 'sofc-test.csv'), rtol=1e-7)
+        self.compare(data, self.test_data_path / "sofc-test.csv", rtol=1e-7)
 
 
 class TestDuplicateReactions(utilities.CanteraTest):
@@ -889,8 +887,7 @@ def test_fromCti(self):
 
     def test_fromXml(self):
         import xml.etree.ElementTree as ET
-        p = os.path.dirname(__file__)
-        root = ET.parse(pjoin(p, '..', 'data', 'h2o2.xml')).getroot()
+        root = ET.parse(self.cantera_data_path / "h2o2.xml").getroot()
         rxn_node = root.find('.//reaction[@id="0001"]')
         r = ct.Reaction.fromXml(ET.tostring(rxn_node))
 
@@ -925,8 +922,7 @@ def test_listFromFile(self):
 
     def test_listFromCti(self):
         gas = ct.Solution("h2o2.xml", transport_model=None)
-        p = os.path.dirname(__file__)
-        with open(pjoin(p, '..', 'data', 'h2o2.cti')) as f:
+        with open(self.cantera_data_path / "h2o2.cti") as f:
             R = ct.Reaction.listFromCti(f.read())
         eq1 = [r.equation for r in R]
         eq2 = [r.equation for r in gas.reactions()]
@@ -934,8 +930,7 @@ def test_listFromCti(self):
 
     def test_listFromXml(self):
         gas = ct.Solution("h2o2.xml", transport_model=None)
-        p = os.path.dirname(__file__)
-        with open(pjoin(p, '..', 'data', 'h2o2.xml')) as f:
+        with open(self.cantera_data_path / "h2o2.xml") as f:
             R = ct.Reaction.listFromXml(f.read())
         eq1 = [r.equation for r in R]
         eq2 = [r.equation for r in gas.reactions()]
diff --git a/interfaces/cython/cantera/test/test_onedim.py b/interfaces/cython/cantera/test/test_onedim.py
index 753ce16a3e5..5cc90278a4e 100644
--- a/interfaces/cython/cantera/test/test_onedim.py
+++ b/interfaces/cython/cantera/test/test_onedim.py
@@ -1,8 +1,6 @@
 import cantera as ct
 from . import utilities
 import numpy as np
-import os
-from os.path import join as pjoin
 
 from cantera.composite import _h5py
 
@@ -510,9 +508,9 @@ def test_save_restore(self):
 
         self.create_sim(p, Tin, reactants)
         self.solve_fixed_T()
-        filename = pjoin(self.test_work_dir, 'onedim-fixed-T.xml')
-        if os.path.exists(filename):
-            os.remove(filename)
+        filename = self.test_work_path / "onedim-fixed-T.xml"
+        if filename.is_file():
+            filename.unlink()
 
         Y1 = self.sim.Y
         u1 = self.sim.velocity
@@ -577,9 +575,9 @@ def test_save_restore_add_species(self):
         p = 2 * ct.one_atm
         Tin = 400
 
-        filename = pjoin(self.test_work_dir, 'onedim-add-species.xml')
-        if os.path.exists(filename):
-            os.remove(filename)
+        filename = self.test_work_path / "onedim-add-species.xml"
+        if filename.is_file():
+            filename.unlink()
 
         self.create_sim(p, Tin, reactants, mech='h2o2.xml')
         gas1 = self.gas
@@ -606,9 +604,9 @@ def test_save_restore_remove_species(self):
         p = 2 * ct.one_atm
         Tin = 400
 
-        filename = pjoin(self.test_work_dir, 'onedim-add-species.xml')
-        if os.path.exists(filename):
-            os.remove(filename)
+        filename = self.test_work_path / "onedim-add-species.xml"
+        if filename.is_file():
+            filename.unlink()
 
         self.create_sim(p, Tin, reactants, mech='h2o2-plus.xml')
         gas1 = self.gas
@@ -630,9 +628,9 @@ def test_save_restore_remove_species(self):
             self.assertArrayNear(Y1[k1], Y2[k2])
 
     def test_write_csv(self):
-        filename = pjoin(self.test_work_dir, 'onedim-write_csv.csv')
-        if os.path.exists(filename):
-            os.remove(filename)
+        filename = self.test_work_path / "onedim-write_csv.csv"
+        if filename.is_file():
+            filename.unlink()
 
         self.create_sim(2e5, 350, 'H2:1.0, O2:2.0', mech='h2o2.xml')
         self.sim.write_csv(filename)
@@ -645,9 +643,9 @@ def test_write_csv(self):
 
     @utilities.unittest.skipIf(isinstance(_h5py, ImportError), "h5py is not installed")
     def test_write_hdf(self):
-        filename = pjoin(self.test_work_dir, 'onedim-write_hdf.h5')
-        if os.path.exists(filename):
-            os.remove(filename)
+        filename = self.test_work_path / "onedim-write_hdf.h5"
+        if filename.is_file():
+            filename.unlink()
 
         self.run_mix(phi=1.1, T=350, width=2.0, p=2.0, refine=False)
         desc = 'mixture-averaged simulation'
@@ -771,7 +769,7 @@ def solve_mix(self, ratio=3.0, slope=0.1, curve=0.12, prune=0.0):
 
     @utilities.slow_test
     def test_mixture_averaged(self, saveReference=False):
-        referenceFile = pjoin(self.test_data_dir, 'DiffusionFlameTest-h2-mix.csv')
+        referenceFile = self.test_data_path / "DiffusionFlameTest-h2-mix.csv"
         self.create_sim(p=ct.one_atm)
 
         nPoints = len(self.sim.grid)
@@ -796,7 +794,7 @@ def test_mixture_averaged(self, saveReference=False):
             self.assertFalse(bad, bad)
 
     def test_auto(self, saveReference=False):
-        referenceFile = pjoin(self.test_data_dir, 'DiffusionFlameTest-h2-auto.csv')
+        referenceFile = self.test_data_path / "DiffusionFlameTest-h2-auto.csv"
         self.create_sim(p=ct.one_atm, mdot_fuel=2, mdot_ox=3)
 
         nPoints = []
@@ -885,7 +883,7 @@ def test_restart(self):
         self.assertNear(mdot[-1], -self.sim.oxidizer_inlet.mdot, 1e-4)
 
     def test_mixture_averaged_rad(self, saveReference=False):
-        referenceFile = pjoin(self.test_data_dir, 'DiffusionFlameTest-h2-mix-rad.csv')
+        referenceFile = self.test_data_path / "DiffusionFlameTest-h2-mix-rad.csv"
         self.create_sim(p=ct.one_atm)
 
         nPoints = len(self.sim.grid)
@@ -916,12 +914,12 @@ def test_mixture_averaged_rad(self, saveReference=False):
                                             rtol=1e-2, atol=1e-8, xtol=1e-2)
             self.assertFalse(bad, bad)
 
-        filename = pjoin(self.test_work_dir, 'DiffusionFlameTest-h2-mix-rad.csv')
+        filename = self.test_work_path / "DiffusionFlameTest-h2-mix-rad.csv"
         self.sim.write_csv(filename) # check output
-        self.assertTrue(os.path.exists(filename))
+        self.assertTrue(filename.is_file())
         csv_data = np.genfromtxt(filename, dtype=float, delimiter=',', names=True)
         self.assertIn('qdot', csv_data.dtype.names)
-        os.remove(filename)
+        filename.unlink()
 
     def test_strain_rate(self):
         # This doesn't test that the values are correct, just that they can be
@@ -1006,7 +1004,7 @@ def test_mixture_averaged(self, saveReference=False):
         data[:,3] = sim.T
         data[:,4:] = sim.Y.T
 
-        referenceFile = pjoin(self.test_data_dir, 'CounterflowPremixedFlame-h2-mix.csv')
+        referenceFile = self.test_data_path / "CounterflowPremixedFlame-h2-mix.csv"
         if saveReference:
             np.savetxt(referenceFile, data, '%11.6e', ', ')
         else:
@@ -1014,12 +1012,12 @@ def test_mixture_averaged(self, saveReference=False):
                                             rtol=1e-2, atol=1e-8, xtol=1e-2)
             self.assertFalse(bad, bad)
 
-        filename = pjoin(self.test_work_dir, 'CounterflowPremixedFlame-h2-mix.csv')
+        filename = self.test_work_path / "CounterflowPremixedFlame-h2-mix.csv"
         sim.write_csv(filename) # check output
-        self.assertTrue(os.path.exists(filename))
+        self.assertTrue(filename.is_file())
         csv_data = np.genfromtxt(filename, dtype=float, delimiter=',', names=True)
         self.assertNotIn('qdot', csv_data.dtype.names)
-        os.remove(filename)
+        filename.unlink()
 
     def run_case(self, phi, T, width, P):
         gas = ct.Solution('h2o2.xml')
@@ -1190,9 +1188,9 @@ def test_reacting_surface_case3(self):
 
     @utilities.unittest.skipIf(isinstance(_h5py, ImportError), "h5py is not installed")
     def test_write_hdf(self):
-        filename = pjoin(self.test_work_dir, 'impingingjet-write_hdf.h5')
-        if os.path.exists(filename):
-            os.remove(filename)
+        filename = self.test_work_path / "impingingjet-write_hdf.h5"
+        if filename.is_file():
+            filename.unlink()
 
         self.run_reacting_surface(xch4=0.095, tsurf=900.0, mdot=0.06, width=0.1)
         self.sim.write_hdf(filename)
diff --git a/interfaces/cython/cantera/test/test_reactor.py b/interfaces/cython/cantera/test/test_reactor.py
index 7ca13a34404..79157887ae8 100644
--- a/interfaces/cython/cantera/test/test_reactor.py
+++ b/interfaces/cython/cantera/test/test_reactor.py
@@ -1,6 +1,6 @@
 import math
 import re
-from os.path import join as pjoin
+from pathlib import Path
 import os
 
 import numpy as np
@@ -1145,8 +1145,8 @@ def test_coverages_regression1(self):
         surf1.coverages = C
         self.assertArrayNear(surf1.coverages, C)
         data = []
-        test_file = pjoin(self.test_work_dir, 'test_coverages_regression1.csv')
-        reference_file = pjoin(self.test_data_dir, 'WallKinetics-coverages-regression1.csv')
+        test_file = self.test_work_path / "test_coverages_regression1.csv"
+        reference_file = self.test_data_path / "WallKinetics-coverages-regression1.csv"
         data = []
         for t in np.linspace(1e-6, 1e-3):
             self.net.advance(t)
@@ -1170,8 +1170,8 @@ def test_coverages_regression2(self):
         surf.coverages = C
         self.assertArrayNear(surf.coverages, C)
         data = []
-        test_file = pjoin(self.test_work_dir, 'test_coverages_regression2.csv')
-        reference_file = pjoin(self.test_data_dir, 'WallKinetics-coverages-regression2.csv')
+        test_file = self.test_work_path / "test_coverages_regression2.csv"
+        reference_file = self.test_data_path / "WallKinetics-coverages-regression2.csv"
         data = []
         for t in np.linspace(1e-6, 1e-3):
             self.net.advance(t)
@@ -1516,7 +1516,7 @@ class CombustorTestImplementation:
     """
 
     def setUp(self):
-        self.referenceFile = pjoin(os.path.dirname(__file__), 'data', 'CombustorTest-integrateWithAdvance.csv')
+        self.referenceFile = Path(__file__).parent / "data" / "CombustorTest-integrateWithAdvance.csv"
         self.gas = ct.Solution('h2o2.yaml', transport_model=None)
 
         # create a reservoir for the fuel inlet, and set to pure methane.
@@ -1622,7 +1622,7 @@ class WallTestImplementation:
     """
 
     def setUp(self):
-        self.referenceFile = pjoin(os.path.dirname(__file__), 'data', 'WallTest-integrateWithAdvance.csv')
+        self.referenceFile = Path(__file__).parent / "data" / "WallTest-integrateWithAdvance.csv"
         # reservoir to represent the environment
         self.gas0 = ct.Solution('air.xml')
         self.gas0.TP = 300, ct.one_atm
diff --git a/interfaces/cython/cantera/test/test_thermo.py b/interfaces/cython/cantera/test/test_thermo.py
index b7b34349b6b..69b2df21eba 100644
--- a/interfaces/cython/cantera/test/test_thermo.py
+++ b/interfaces/cython/cantera/test/test_thermo.py
@@ -1,7 +1,4 @@
-from os.path import join as pjoin
-from pathlib import Path
 from collections import OrderedDict
-import os
 import numpy as np
 import gc
 import warnings
@@ -1113,8 +1110,7 @@ def test_fromCti(self):
 
     def test_fromXml(self):
         import xml.etree.ElementTree as ET
-        p = os.path.dirname(__file__)
-        root = ET.parse(pjoin(p, '..', 'data', 'h2o2.xml')).getroot()
+        root = ET.parse(str(self.cantera_data_path / "h2o2.xml")).getroot()
         h2_node = root.find('.//species[@name="H2"]')
         h2_string = ET.tostring(h2_node)
 
@@ -1138,8 +1134,7 @@ def test_listfromFile_yaml(self):
         self.assertEqual({sp.name for sp in S}, set(self.gas.species_names))
 
     def test_listFromCti(self):
-        p = os.path.dirname(__file__)
-        with open(pjoin(p, '..', 'data', 'h2o2.cti')) as f:
+        with open(self.cantera_data_path / "h2o2.cti") as f:
             S = ct.Species.listFromCti(f.read())
 
         self.assertEqual(S[3].name, self.gas.species_name(3))
@@ -1160,15 +1155,14 @@ def test_listFomYaml(self):
 
     def test_listFromYaml_section(self):
         species = ct.Species.listFromYaml(
-            Path(__file__).parent.joinpath('data', 'ideal-gas.yaml').read_text(),
+            (self.test_data_path / "ideal-gas.yaml").read_text(),
             'species')
 
         self.assertEqual(species[0].name, 'O2')
         self.assertEqual(species[1].composition, {'N': 1, 'O': 1})
 
     def test_listFromXml(self):
-        p = os.path.dirname(__file__)
-        with open(pjoin(p, '..', 'data', 'h2o2.xml')) as f:
+        with open(self.cantera_data_path / "h2o2.xml") as f:
             S = ct.Species.listFromXml(f.read())
 
         self.assertEqual(S[4].name, self.gas.species_name(4))
diff --git a/interfaces/cython/cantera/test/utilities.py b/interfaces/cython/cantera/test/utilities.py
index 34c47a3da91..583611ac923 100644
--- a/interfaces/cython/cantera/test/utilities.py
+++ b/interfaces/cython/cantera/test/utilities.py
@@ -6,6 +6,7 @@
 import tempfile
 import unittest
 import errno
+from pathlib import Path
 
 try:
     import ruamel_yaml as yaml
@@ -31,36 +32,40 @@ def setUpClass(cls):
         # an in-source test, create the directory in the root
         # test/work directory. Otherwise, create a system level
         # temporary directory
-        root_dir = os.path.abspath(os.path.join(os.path.dirname(__file__),
-                                                '..', '..', '..', '..'))
-        if os.path.exists(os.path.join(root_dir, 'SConstruct')):
-            cls.test_work_dir = os.path.join(root_dir, 'test', 'work', 'python')
+        root_dir = Path(__file__).parents[4].resolve()
+        if (root_dir / "Sconstruct").is_file():
+            cls.test_work_path = root_dir / "test" / "work" / "python"
             cls.using_tempfile = False
             try:
-                os.makedirs(cls.test_work_dir)
-            except OSError as e:
-                if e.errno == errno.EEXIST:
-                    pass
-                elif e.errno == errno.EACCES:
-                    cls.test_work_dir = tempfile.mkdtemp()
-                else:
-                    raise
+                cls.test_work_path.mkdir()
+            except FileExistsError:
+                pass
+            except FileNotFoundError:
+                cls.test_work_path = Path(tempfile.mkdtemp())
+                cls.using_tempfile = True
+            except:
+                raise
         else:
-            cls.test_work_dir = tempfile.mkdtemp()
+            cls.test_work_path = Path(tempfile.mkdtemp())
             cls.using_tempfile = True
 
         cantera.make_deprecation_warnings_fatal()
-        cantera.add_directory(cls.test_work_dir)
-        cls.test_data_dir = os.path.abspath(os.path.join(os.path.dirname(__file__), 'data'))
-        cls.cantera_data = os.path.abspath(os.path.join(
-            os.path.dirname(__file__), '..', 'data'))
+        cantera.add_directory(cls.test_work_path)
+        cls.test_data_path = Path(__file__).parent / "data"
+        cls.cantera_data_path = Path(__file__).parents[1] / "data"
+
+        # retain old path strings
+        cls.test_work_dir = str(cls.test_work_path)
+        cls.test_data_dir = str(cls.test_data_path)
+        cls.cantera_data = str(cls.cantera_data_path)
+
 
     @classmethod
     def tearDownClass(cls):
         # Remove the working directory after testing, but only if its a temp directory
         if getattr(cls, "using_tempfile", False):
             try:
-                shutil.rmtree(cls.test_work_dir)
+                shutil.rmtree(str(cls.test_work_path))
             except OSError:
                 pass
 
@@ -101,7 +106,7 @@ def compare(self, data, reference_file, rtol=1e-8, atol=1e-12):
         file if it does not exist.
         """
         data = np.array(data)
-        if os.path.exists(reference_file):
+        if Path(reference_file).is_file():
             # Compare with existing output file
             ref = np.genfromtxt(reference_file)
             self.assertEqual(data.shape, ref.shape)
@@ -110,7 +115,7 @@ def compare(self, data, reference_file, rtol=1e-8, atol=1e-12):
         else:
             # Generate the output file for the first time
             warnings.warn('Generating test data file:' +
-                          os.path.abspath(reference_file))
+                          Path(reference_file).resolve())
             np.savetxt(reference_file, data, fmt='%.10e')
 
 
@@ -131,12 +136,12 @@ def compareProfiles(reference, sample, rtol=1e-5, atol=1e-12, xtol=1e-5):
     If the comparison succeeds, this function returns `None`. If the comparison
     fails, a formatted report of the differing elements is returned.
     """
-    if isinstance(reference, str):
+    if isinstance(reference, (str, Path)):
         reference = np.genfromtxt(reference, delimiter=',').T
     else:
         reference = np.asarray(reference).T
 
-    if isinstance(sample, str):
+    if isinstance(sample, (str, Path)):
         sample = np.genfromtxt(sample, delimiter=',').T
     else:
         sample = np.asarray(sample).T