diff --git a/pyfeeds/evaluate.py b/pyfeeds/evaluate.py
index 287a060d613fec7bc5a00a4339954860cb6475af..e4eb575d239a875357b6bc0860069d51809e4384 100644
--- a/pyfeeds/evaluate.py
+++ b/pyfeeds/evaluate.py
@@ -420,9 +420,11 @@ def loadImage(pyf, filename):
         return img, data
 
 
-def evalHeader(testfile, benchmark, alldims=True, pyf=None):
+def evalHeader(testfile, benchmark, ndims=None, pyf=None):
     """Evaluation routine which compares the header fields of two NIFTI
-    images.
+    images. By default the dim/pixdim values for every dimension are compared,
+    but the ``ndims`` argument can be used if you only want to compare the
+    first three dimensions for instance.
 
     Returns 0 if they all match, 1 otherwise.
     """
@@ -437,14 +439,17 @@ def evalHeader(testfile, benchmark, alldims=True, pyf=None):
               'qoffset_x', 'qoffset_y',  'qoffset_z',
               'srow_x',    'srow_y',     'srow_z']
 
+    if ndims is None:
+        ndims = max(img1.header['dim'][0],
+                    img2.header['dim'][0])
+
     for f in fields:
         f1 = hdr1[f]
         f2 = hdr2[f]
 
-        if (not alldims) and (f in ('dim', 'pixdim')):
-            ndim = img1.header['dim'][0]
-            f1   = f1[:ndim + 1]
-            f2   = f2[:ndim + 1]
+        if f in ('dim', 'pixdim'):
+            f1 = f1[:ndims + 1]
+            f2 = f2[:ndims + 1]
 
         if not np.all(np.isclose(f1, f2)):
             return 1
@@ -453,14 +458,8 @@ def evalHeader(testfile, benchmark, alldims=True, pyf=None):
 
 
 def evalHeaderRestrictDims(testfile, benchmark, pyf=None):
-    """Evaluation routine which compares the header fields of two NIFTI
-    images. For the `dim` and `pixdim` fields, only the entries which
-    are expected to be valid (e.g. `dim1`, `dim2`, and `dim3` for a 3D image)
-    are compared.
-
-    Returns 0 if they all match, 1 otherwise.
-    """
-    return evalHeader(testfile, benchmark, alldims=False, pyf=pyf)
+    """Legacy alias for ``evalHeader``. """
+    return evalHeader(testfile, benchmark, pyf=pyf)
 
 
 def evalImage(testfile, benchmark, pyf=None):