diff --git a/tensorforce/core/explorations/__init__.py b/tensorforce/core/explorations/__init__.py
index f1668623..420f8d4a 100755
--- a/tensorforce/core/explorations/__init__.py
+++ b/tensorforce/core/explorations/__init__.py
@@ -17,6 +17,7 @@
 from tensorforce.core.explorations.exploration import Exploration
 from tensorforce.core.explorations.constant import Constant
 from tensorforce.core.explorations.linear_decay import LinearDecay
+from tensorforce.core.explorations.epsilon_anneal import EpsilonAnneal
 from tensorforce.core.explorations.epsilon_decay import EpsilonDecay
 from tensorforce.core.explorations.ornstein_uhlenbeck_process import OrnsteinUhlenbeckProcess
 
@@ -24,6 +25,7 @@
 explorations = dict(
     constant=Constant,
     linear_decay=LinearDecay,
+    epsilon_anneal=EpsilonAnneal,
     epsilon_decay=EpsilonDecay,
     ornstein_uhlenbeck=OrnsteinUhlenbeckProcess
 )
diff --git a/tensorforce/core/explorations/epsilon_anneal.py b/tensorforce/core/explorations/epsilon_anneal.py
new file mode 100644
index 00000000..e8af713e
--- /dev/null
+++ b/tensorforce/core/explorations/epsilon_anneal.py
@@ -0,0 +1,37 @@
+# Copyright 2017 reinforce.io. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ==============================================================================
+
+from tensorforce.core.explorations import Exploration
+
+
+class EpsilonAnneal(Exploration):
+    """
+    Annealing epsilon parameter based on ratio of current timestep to total timesteps.
+    """
+
+    def __init__(self, epsilon=1.0, epsilon_final=0.1, epsilon_timesteps=10000):
+        self.epsilon = epsilon
+        self.epsilon_final = epsilon_final
+        self.epsilon_timesteps = epsilon_timesteps
+
+    def __call__(self, episode=0, timestep=0):
+        # TODO: Trim by length of `first_update`, removing steps with no learning.
+        offset = 0 # self.first_update
+        self.epsilon = min(1.0, max(
+            self.epsilon_final,
+            1.0-(timestep - offset)/(self.epsilon_timesteps - offset)
+        ));
+
+        return self.epsilon
diff --git a/tensorforce/core/explorations/epsilon_decay.py b/tensorforce/core/explorations/epsilon_decay.py
index fb70bd5b..5d184251 100755
--- a/tensorforce/core/explorations/epsilon_decay.py
+++ b/tensorforce/core/explorations/epsilon_decay.py
@@ -18,8 +18,8 @@
 
 class EpsilonDecay(Exploration):
     """
-    Linearly decaying epsilon parameter based on number of states,
-    an initial random epsilon and a final random epsilon.
+    Exponentially decaying epsilon parameter based on ratio of
+    difference between current and final epsilon to total timesteps.
     """
 
     def __init__(self, epsilon=1.0, epsilon_final=0.1, epsilon_timesteps=10000):