diff --git a/doc/source/v0.14.1.txt b/doc/source/v0.14.1.txt
index c1e5877d09004..ad2e4ad44b98f 100644
--- a/doc/source/v0.14.1.txt
+++ b/doc/source/v0.14.1.txt
@@ -249,3 +249,5 @@ Bug Fixes
 
 - Bug in non-monotonic ``Index.union`` may preserve ``name`` incorrectly (:issue:`7458`)
 - Bug in ``DatetimeIndex.intersection`` doesn't preserve timezone (:issue:`4690`)
+- Bug in ``Float64Index`` slicing with ``__getitem__`` (``[]``) style indexing
+  (:issue:`7501`).
diff --git a/pandas/core/common.py b/pandas/core/common.py
index 92d60ae8d8847..c29016fbe39e7 100644
--- a/pandas/core/common.py
+++ b/pandas/core/common.py
@@ -8,6 +8,7 @@
 import codecs
 import csv
 import types
+from functools import partial
 from datetime import datetime, timedelta
 
 from numpy.lib.format import read_array, write_array
@@ -19,6 +20,7 @@
 import pandas.tslib as tslib
 from pandas import compat
 from pandas.compat import StringIO, BytesIO, range, long, u, zip, map
+import pandas.tools as tools
 
 from pandas.core.config import get_option
 from pandas.core import array as pa
@@ -2290,6 +2292,18 @@ def _is_sequence(x):
         return False
 
 
+def is_slice_of(typ_validator):
+    int_or_none = lambda x: x is None or typ_validator(x)
+    attrs = 'start', 'stop', 'step'
+    return lambda slc: all(map(tools.util.compose(int_or_none,
+                                                  partial(getattr, slc)),
+                               attrs))
+
+
+is_int_slice = is_slice_of(is_integer)
+
+
+
 _ensure_float64 = algos.ensure_float64
 _ensure_float32 = algos.ensure_float32
 _ensure_int64 = algos.ensure_int64
diff --git a/pandas/core/index.py b/pandas/core/index.py
index 5956cd7eea9c0..1de588f85c4de 100644
--- a/pandas/core/index.py
+++ b/pandas/core/index.py
@@ -10,6 +10,7 @@
 import pandas.lib as lib
 import pandas.algos as _algos
 import pandas.index as _index
+import pandas.tools as tools
 from pandas.lib import Timestamp, is_datetime_array
 from pandas.core.base import FrozenList, FrozenNDArray, IndexOpsMixin
 from pandas.util.decorators import cache_readonly, deprecate
@@ -2004,13 +2005,10 @@ def _convert_scalar_indexer(self, key, typ=None):
     def _convert_slice_indexer(self, key, typ=None):
         """ convert a slice indexer, by definition these are labels
             unless we are iloc """
-        if typ == 'iloc':
+        if typ == 'iloc' or (typ == 'getitem' and com.is_int_slice(key)):
             return super(Float64Index, self)._convert_slice_indexer(key,
                                                                     typ=typ)
-
-        # allow floats here
-        validator = lambda v: v is None or is_integer(v) or is_float(v)
-        self._validate_slicer(key, validator)
+        self._validate_slicer(key, lambda v: v is None or is_float(v))
 
         # translate to locations
         return self.slice_indexer(key.start, key.stop, key.step)
diff --git a/pandas/tests/test_indexing.py b/pandas/tests/test_indexing.py
index c074c4333a774..d2ebd17f02853 100644
--- a/pandas/tests/test_indexing.py
+++ b/pandas/tests/test_indexing.py
@@ -3629,6 +3629,16 @@ def test_duplicate_ix_returns_series(self):
         e = df.loc[0.2, 'a']
         tm.assert_series_equal(r, e)
 
+    def test_float_slice_ix_style(self):
+        n = 5
+        df = DataFrame(np.random.randn(n, 2), columns=list('ab'),
+                       index=np.arange(0.0, n) + 50)
+        expected = df.iloc[2:4]
+        result = df[2:4]
+        tm.assert_frame_equal(expected, result)
+
+        assert df[2.0:4.0].empty
+
 
 if __name__ == '__main__':
     nose.runmodule(argv=[__file__, '-vvs', '-x', '--pdb', '--pdb-failure'],