diff --git a/.github/workflows/codecov.yml b/.github/workflows/codecov.yml
index d7ec1cf..7383136 100644
--- a/.github/workflows/codecov.yml
+++ b/.github/workflows/codecov.yml
@@ -1,27 +1,37 @@
-name: Python application
+name: Python application cross-platform test
 
-on: [push, pull_request]
+on: [push]
 
 jobs:
   build:
+    runs-on: ${{ matrix.os }}
 
-    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        os: [ubuntu-latest, windows-latest, macos-latest]
+        python-version: [3.8, 3.9, "3.10", "3.11", "3.12"]
 
     steps:
       - uses: actions/checkout@v2
-      - name: Set up Python 3.10
+
+      - name: Set up Python ${{ matrix.python-version }}
         uses: actions/setup-python@v2
         with:
-          python-version: "3.10"
+          python-version: ${{ matrix.python-version }}
+
       - name: Install dependencies
         run: |
           python -m pip install --upgrade pip
-          pip install pytest pytest-cov coverage
-          if [ -f requirements.txt ]; then pip install -r requirements.txt; fi
+          pip install -r requirements-test.txt
+          pip install -r requirements.txt
+
       - name: Test with pytest
         run: |
-          pytest --cov=brokenaxes --cov-report=xml test.py
+          pytest --cov=brokenaxes --cov-report=xml --mpl --mpl-baseline-path test_baseline test.py
+
       - name: Upload coverage to Codecov
         uses: codecov/codecov-action@v3
-        env:
-          CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
+        if: matrix.python-version == '3.10' && matrix.os == 'ubuntu-latest'
+        with:
+          token: ${{ secrets.CODECOV_TOKEN }}
+          fail_ci_if_error: true
diff --git a/requirements-test.txt b/requirements-test.txt
new file mode 100644
index 0000000..a2e2b0a
--- /dev/null
+++ b/requirements-test.txt
@@ -0,0 +1,3 @@
+pytest>=6.0
+pytest-mpl==0.17.0
+pytest-cov
\ No newline at end of file
diff --git a/setup.py b/setup.py
index ce01cfb..ea89268 100644
--- a/setup.py
+++ b/setup.py
@@ -1,4 +1,4 @@
-from setuptools import setup, find_packages
+from setuptools import setup
 # To use a consistent encoding
 from codecs import open
 from os import path
@@ -24,14 +24,14 @@
         'Topic :: Scientific/Engineering',
         'Framework :: Matplotlib',
         'License :: OSI Approved :: MIT License',
-        'Programming Language :: Python :: 3.5',
-        'Programming Language :: Python :: 3.6',
-        'Programming Language :: Python :: 3.7'
+        'Programming Language :: Python :: 3.8',
+        'Programming Language :: Python :: 3.9',
+        'Programming Language :: Python :: 3.10',
+        'Programming Language :: Python :: 3.11',
+        'Programming Language :: Python :: 3.12',
     ],
 
     keywords='data visualization',
-
-    #packages=find_packages(exclude=['docs']),
     py_modules=["brokenaxes"],
 
     # List run-time dependencies here.  These will be installed by pip when
@@ -43,10 +43,9 @@
     # List additional groups of dependencies here (e.g. development
     # dependencies). You can install these using the following syntax,
     # for example:
-    # $ pip install -e .[dev,test]
+    # $ pip install -e .[test]
     extras_require={
-    #    'dev': ['check-manifest'],
-        'test': ['pytest'],
+        'test': ['pytest>=6.0', 'pytest-cov', 'pytest-mpl'],
     },
 
 
diff --git a/test.py b/test.py
index 4390389..af77ab3 100644
--- a/test.py
+++ b/test.py
@@ -9,6 +9,10 @@
 from brokenaxes import brokenaxes
 
 
+np.random.seed(42)
+
+
+@pytest.mark.mpl_image_compare
 def test_standard():
     fig = plt.figure(figsize=(5, 2))
     bax = brokenaxes(
@@ -20,8 +24,10 @@ def test_standard():
     bax.legend(loc=3)
     bax.set_xlabel("time")
     bax.set_ylabel("value")
+    return fig
 
 
+@pytest.mark.mpl_image_compare
 def test_subplots():
     sps1, sps2 = GridSpec(2, 1)
 
@@ -33,7 +39,10 @@ def test_subplots():
     bax = brokenaxes(xlims=((0, 2.5), (3, 6)), subplot_spec=sps2)
     bax.hist(x, histtype="bar")
 
+    return bax.fig
 
+
+@pytest.mark.mpl_image_compare
 def test_log():
     fig = plt.figure(figsize=(5, 5))
     bax = brokenaxes(
@@ -53,7 +62,10 @@ def test_log():
     bax.set_xlabel("x")
     bax.set_ylabel("y")
 
+    return fig
+
 
+@pytest.mark.mpl_image_compare
 def test_datetime():
     fig = plt.figure(figsize=(5, 5))
     xx = [datetime.datetime(2020, 1, x) for x in range(1, 20)]
@@ -79,7 +91,10 @@ def test_datetime():
     [x.remove() for x in bax.diag_handles]
     bax.draw_diags()
 
+    return fig
 
+
+@pytest.mark.mpl_image_compare
 def test_datetime_y():
     fig = plt.figure(figsize=(5, 5))
     yy = [datetime.datetime(2020, 1, x) for x in range(1, 20)]
@@ -103,7 +118,10 @@ def test_datetime_y():
 
     bax.draw_diags()
 
+    return fig
+
 
+@pytest.mark.mpl_image_compare
 def test_legend():
     fig = plt.figure(figsize=(5, 2))
     bax = brokenaxes(
@@ -114,27 +132,28 @@ def test_legend():
     h2 = bax.plot(x, np.cos(10 * x), label="cos")
     bax.legend(handles=[h1[0][0], h2[0][0]], labels=["1", "2"])
 
+    return fig
 
+
+@pytest.mark.mpl_image_compare
 def test_text():
     bax = brokenaxes(
         xlims=((0, 0.1), (0.4, 0.7)), ylims=((-1, 0.7), (0.79, 1)), hspace=0.05
     )
     bax.text(0.5, 0.5, "hello")
 
-
-def test_text_error():
-    bax = brokenaxes(
-        xlims=((0, 0.1), (0.4, 0.7)), ylims=((-1, 0.7), (0.79, 1)), hspace=0.05
-    )
-    with pytest.raises(ValueError):
-        bax.text(-11, -11, "hello")
+    return bax.fig
 
 
+@pytest.mark.mpl_image_compare
 def test_lims_arrays():
     lims = np.arange(6).reshape((-1,2))
     brokenaxes(xlims=lims, ylims=lims)
 
+    return plt.gcf()
+
 
+@pytest.mark.mpl_image_compare
 def test_pass_fig():
     fig = plt.figure(figsize=(5, 2))
     bax = brokenaxes(
@@ -142,7 +161,10 @@ def test_pass_fig():
     )
     assert bax.fig is fig
 
+    return fig
 
+
+@pytest.mark.mpl_image_compare
 def test_despine():
     fig = plt.figure(figsize=(5, 2))
     bax = brokenaxes(
@@ -150,7 +172,10 @@ def test_despine():
     )
     assert bax.despine is False
 
+    return fig
+
 
+@pytest.mark.mpl_image_compare
 def test_set_title():
     fig = plt.figure(figsize=(5, 2))
     bax = brokenaxes(
@@ -158,7 +183,10 @@ def test_set_title():
     )
     bax.set_title("title")
 
+    return fig
 
+
+@pytest.mark.mpl_image_compare
 def test_secondary_axes():
 
     fig = plt.figure(figsize=(5, 2))
@@ -171,6 +199,19 @@ def test_secondary_axes():
     bax.secondary_yaxis("left", label="left")
     bax.secondary_yaxis("right")
 
+    return fig
+
+
+@pytest.mark.mpl_image_compare
+def test_draw_diags():
+    fig = plt.figure(figsize=(5, 2))
+    bax = brokenaxes(
+        xlims=((0, 0.1), (0.4, 0.7)), ylims=((-1, 0.7), (0.79, 1)), hspace=0.05
+    )
+    bax.draw_diags(tilt=90, d=.05)
+
+    return fig
+
 
 def test_get_axis_special():
     fig = plt.figure(figsize=(5, 2))
@@ -183,9 +224,9 @@ def test_get_axis_special():
     assert isinstance(bax.get_shared_y_axes(), mpl.cbook.GrouperView)
 
 
-def test_draw_diags():
-    fig = plt.figure(figsize=(5, 2))
+def test_text_error():
     bax = brokenaxes(
         xlims=((0, 0.1), (0.4, 0.7)), ylims=((-1, 0.7), (0.79, 1)), hspace=0.05
     )
-    bax.draw_diags(tilt=90, d=.05)
+    with pytest.raises(ValueError):
+        bax.text(-11, -11, "hello")
\ No newline at end of file
diff --git a/test_baseline/test_datetime.png b/test_baseline/test_datetime.png
new file mode 100644
index 0000000..8455926
Binary files /dev/null and b/test_baseline/test_datetime.png differ
diff --git a/test_baseline/test_datetime_y.png b/test_baseline/test_datetime_y.png
new file mode 100644
index 0000000..68c9e43
Binary files /dev/null and b/test_baseline/test_datetime_y.png differ
diff --git a/test_baseline/test_despine.png b/test_baseline/test_despine.png
new file mode 100644
index 0000000..8835f8b
Binary files /dev/null and b/test_baseline/test_despine.png differ
diff --git a/test_baseline/test_draw_diags.png b/test_baseline/test_draw_diags.png
new file mode 100644
index 0000000..2a17bd5
Binary files /dev/null and b/test_baseline/test_draw_diags.png differ
diff --git a/test_baseline/test_legend.png b/test_baseline/test_legend.png
new file mode 100644
index 0000000..4d2d01e
Binary files /dev/null and b/test_baseline/test_legend.png differ
diff --git a/test_baseline/test_lims_arrays.png b/test_baseline/test_lims_arrays.png
new file mode 100644
index 0000000..7fc586e
Binary files /dev/null and b/test_baseline/test_lims_arrays.png differ
diff --git a/test_baseline/test_log.png b/test_baseline/test_log.png
new file mode 100644
index 0000000..5d377a7
Binary files /dev/null and b/test_baseline/test_log.png differ
diff --git a/test_baseline/test_pass_fig.png b/test_baseline/test_pass_fig.png
new file mode 100644
index 0000000..13df808
Binary files /dev/null and b/test_baseline/test_pass_fig.png differ
diff --git a/test_baseline/test_secondary_axes.png b/test_baseline/test_secondary_axes.png
new file mode 100644
index 0000000..c874a14
Binary files /dev/null and b/test_baseline/test_secondary_axes.png differ
diff --git a/test_baseline/test_set_title.png b/test_baseline/test_set_title.png
new file mode 100644
index 0000000..c428048
Binary files /dev/null and b/test_baseline/test_set_title.png differ
diff --git a/test_baseline/test_standard.png b/test_baseline/test_standard.png
new file mode 100644
index 0000000..c443100
Binary files /dev/null and b/test_baseline/test_standard.png differ
diff --git a/test_baseline/test_subplots.png b/test_baseline/test_subplots.png
new file mode 100644
index 0000000..2105aef
Binary files /dev/null and b/test_baseline/test_subplots.png differ
diff --git a/test_baseline/test_text.png b/test_baseline/test_text.png
new file mode 100644
index 0000000..4adba1b
Binary files /dev/null and b/test_baseline/test_text.png differ