diff --git a/ci/requirements-2.7.sh b/ci/requirements-2.7.sh
index 64d470e5c6e0ed..96a83c083f735b 100644
--- a/ci/requirements-2.7.sh
+++ b/ci/requirements-2.7.sh
@@ -2,6 +2,6 @@
 
 source activate pandas
 
-echo "install 27"
+echo "install 27 BUILD_TEST"
 
 conda install -n pandas -c conda-forge feather-format
diff --git a/ci/requirements-2.7_BUILD_TEST.pip b/ci/requirements-2.7_BUILD_TEST.pip
new file mode 100644
index 00000000000000..59022d9b560051
--- /dev/null
+++ b/ci/requirements-2.7_BUILD_TEST.pip
@@ -0,0 +1,8 @@
+xarray
+geopandas
+dask
+seaborn
+pandas_gbq
+pandas_datareader
+statsmodels
+scikit-learn
diff --git a/ci/requirements-2.7_BUILD_TEST.sh b/ci/requirements-2.7_BUILD_TEST.sh
new file mode 100755
index 00000000000000..d6283a4453f22e
--- /dev/null
+++ b/ci/requirements-2.7_BUILD_TEST.sh
@@ -0,0 +1,7 @@
+#!/bin/bash
+
+source activate pandas
+
+echo "install 27"
+
+conda install -n pandas -c conda-forge pyarrow
diff --git a/pandas/tests/test_downstream.py b/pandas/tests/test_downstream.py
new file mode 100644
index 00000000000000..16fcd9d502fc25
--- /dev/null
+++ b/pandas/tests/test_downstream.py
@@ -0,0 +1,83 @@
+"""
+Testing that we work in the downstream packages
+"""
+import pytest
+from pandas import DataFrame
+from pandas.util import testing as tm
+
+
+@pytest.fixture
+def df():
+    return DataFrame({'A': [1, 2, 3]})
+
+
+def test_dask(df):
+
+    dask = pytest.importorskip('dask')  # noqa
+
+    import dask.dataframe as dd
+
+    ddf = dd.from_pandas(df)
+    assert ddf.A is not None
+    assert ddf.compute()
+
+
+def test_xarray(df):
+
+    xarray = pytest.importorskip('xarray')  # noqa
+
+    assert df.to_xarray() is not None
+
+
+def test_statsmodels():
+
+    statsmodels = pytest.importorskip('statsmodels')  # noqa
+    import statsmodels.api as sm
+    import statsmodels.formula.api as smf
+    df = sm.datasets.get_rdataset("Guerry", "HistData").data
+    smf.ols('Lottery ~ Literacy + np.log(Pop1831)', data=df).fit()
+
+
+def test_scikit_learn(df):
+
+    sklearn = pytest.importorskip('sklearn')  # noqa
+    from sklearn import svm, datasets
+
+    digits = datasets.load_digits()
+    clf = svm.SVC(gamma=0.001, C=100.)
+    clf.fit(digits.data[:-1], digits.target[:-1])
+    clf.predict(digits.data[-1:])
+
+
+def test_seaborn():
+
+    seaborn = pytest.importorskip('seaborn')
+    tips = seaborn.load_dataset("tips")
+    seaborn.stripplot(x="day", y="total_bill", data=tips)
+
+
+def test_pandas_gbq(df):
+
+    pandas_gbq = pytest.importorskip('pandas-gbq')  # noqa
+
+
+@tm.network
+def test_pandas_datareader():
+
+    pandas_datareader = pytest.importorskip('pandas-datareader')  # noqa
+    pandas_datareader.get_data_yahoo('AAPL')
+
+
+def test_geopandas():
+
+    geopandas = pytest.importorskip('geopandas')  # noqa
+    fp = geopandas.datasets.get_path('naturalearth_lowres')
+    assert geopandas.read_file(fp) is not None
+
+
+def test_pyarrow(df):
+
+    pyarrow = pytest.importorskip('pyarrow')  # noqa
+    table = pyarrow.Table.from_pandas(df)
+    result = table.to_pandas()
+    tm.assert_frame_equal(result, df)