diff --git a/skimage/feature/template.py b/skimage/feature/template.py
index b328131f..6ad45926 100644
--- a/skimage/feature/template.py
+++ b/skimage/feature/template.py
@@ -6,12 +6,11 @@ import _template
 from skimage.util.dtype import _convert
 
 
-def match_template(image, template, pad_output=False):
+def match_template(image, template, pad_input=False):
     """Match a template to an image using normalized correlation.
 
     The output is an array with values between -1.0 and 1.0, which correspond
-    to the probability that the template's *origin* (i.e. its top-left
-    corner) is found at that position.
+    to the probability that the template is found at that position.
 
     Parameters
     ----------
@@ -19,18 +18,19 @@ def match_template(image, template, pad_output=False):
         Image to process.
     template : array_like
         Template to locate.
-    pad_output : bool
-        If True, pad output array to be the same size as the input image.
+    pad_input : bool
+        If True, pad `image` with image mean so that output is the same size as
+        the image, and output values correspond to the template center.
         Otherwise, the output is an array with shape `(M - m + 1, N - n + 1)`
-        for an `(M, N)` image and an `(m, n)` template.
+        for an `(M, N)` image and an `(m, n)` template, and matches correspond
+        to origin (top-left corner) of the template.
 
     Returns
     -------
     output : ndarray
-        Correlation results between -1.0 and 1.0. The `output` is truncated
-        (`pad_output = False`) or zero-padded (`pad_output = True`) at the
-        bottom and right edges, where the template would otherwise extend
-        beyond the image edges.
+        Correlation results between -1.0 and 1.0. For an `(M, N)` image and an
+        `(m, n)` template, the `output` is `(M - m + 1, N - n + 1)` when
+        `pad_input = False` and `(M, N)` when `pad_input = True`.
 
     """
     if np.any(np.less(image.shape, template.shape)):
@@ -38,7 +38,7 @@ def match_template(image, template, pad_output=False):
     image = _convert(image, np.float32)
     template = _convert(template, np.float32)
 
-    if pad_output:
+    if pad_input:
         pad_size = tuple(np.array(image.shape) + np.array(template.shape) - 1)
         pad_image = np.mean(image) * np.ones(pad_size, dtype=np.float32)
         h, w = image.shape