diff --git a/interfaces/cython/cantera/test/test_convert.py b/interfaces/cython/cantera/test/test_convert.py
index 858258602..c04b2664e 100644
--- a/interfaces/cython/cantera/test/test_convert.py
+++ b/interfaces/cython/cantera/test/test_convert.py
@@ -2,7 +2,7 @@ import os
 import numpy as np
 import itertools
 
-import utilities
+from . import utilities
 import cantera as ct
 from cantera import ck2cti
 
diff --git a/test/python/runTests.py b/test/python/runTests.py
deleted file mode 100644
index a3609d821..000000000
--- a/test/python/runTests.py
+++ /dev/null
@@ -1,24 +0,0 @@
-"""
-Unit tests for the Cantera Python module.
-
-This script gathers all the tests defined in all of the test<foo>.py
-files, runs them, and prints a report.
-"""
-import sys
-import os
-import unittest
-
-sys.path.insert(0, os.path.abspath('../../interfaces/python'))
-
-import Cantera
-
-Cantera.addDirectory(os.path.join(os.path.split(os.getcwd())[0], 'data'))
-
-if __name__ == '__main__':
-    print '\n* INFO: using Cantera module found at this location:'
-    print '*     ', repr(Cantera.__file__), '\n'
-    sys.stdout.flush()
-
-    loader = unittest.TestLoader()
-    runner = unittest.TextTestRunner(verbosity=2)
-    sys.exit(0)
diff --git a/test/python/utilities.py b/test/python/utilities.py
deleted file mode 100644
index 34c608a5e..000000000
--- a/test/python/utilities.py
+++ /dev/null
@@ -1,62 +0,0 @@
-import numpy as np
-import unittest
-
-class CanteraTest(unittest.TestCase):
-    def assertNear(self, a, b, rtol=1e-8, atol=1e-12, msg=None):
-        cmp = 2 * abs(a - b)/(abs(a) + abs(b) + atol)
-        if cmp > rtol:
-            message = ('AssertNear: %.14g - %.14g = %.14g\n' % (a, b, a-b) +
-                       'Relative error of %10e exceeds rtol = %10e' % (cmp, rtol))
-            if msg:
-                message = msg + '\n' + message
-            self.fail(message)
-
-def compareTimeSeries(reference, sample, rtol=1e-5, atol=1e-12):
-    """
-    Compare two 2D arrays of time series data. Each data set should contain
-    the time in the first column and data series to be compared in successive
-    columns.
-
-    The times in each data set do not need to be the same: The data from the
-    second data set will be interpolated onto the times in the first data set
-    before being compared. This means that the time range of the "sample" data
-    set should be at least as long as the "reference" data set.
-
-    After interpolation, each data point must satisfy either the relative
-    error tolerance specified by `rtol` or the absolute error tolerance
-    specified by `atol`.
-
-    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):
-        reference = np.genfromtxt(reference, delimiter=',').T
-    else:
-        reference = np.asarray(reference).T
-
-    sample = np.asarray(sample).T
-
-    assert reference.shape[0] == sample.shape[0]
-
-    nVars = reference.shape[0]
-    nTimes = reference.shape[1]
-
-    bad = []
-    template = '{0:9.4e}  {1: 3d}   {2:14.8e}  {3:14.8e}  {4:9.3e}  {5:9.3e}'
-    for i in range(1, nVars):
-        comp = np.interp(reference[0], sample[0], sample[i])
-        for j in range(nTimes):
-            a = reference[i,j]
-            b = comp[j]
-            abserr = abs(a-b)
-            relerr = 2 * abs(a-b) / (abs(a) + abs(b) + atol)
-            if abserr > atol and relerr > rtol:
-                bad.append(template.format(reference[0][j], i, a, b, abserr, relerr))
-
-    if bad:
-        header = ['Failed time series comparisons:',
-                  '  time   component  reference       test value    abs. err.  rel. err',
-                  '----------  ---   --------------  --------------  ---------  ---------']
-        return '\n'.join(header + bad)
-    else:
-        return None