diff --git a/lib/galaxy/datatypes/interval.py b/lib/galaxy/datatypes/interval.py
index 89d771b42e6..20728b5d3d8 100644
--- a/lib/galaxy/datatypes/interval.py
+++ b/lib/galaxy/datatypes/interval.py
@@ -477,9 +477,9 @@ class Bed(Interval):
         if not get_headers(file_prefix, '\t', comment_designator='#', count=1):
             return False
         try:
-            headers = iter_headers(file_prefix, '\t', comment_designator='#')
-            for hdr in headers:
-                if hdr[0] == '':
+            found_valid_lines = False
+            for hdr in iter_headers(file_prefix, '\t', comment_designator='#'):
+                if not hdr or hdr == ['']:
                     continue
                 if len(hdr) < 3 or len(hdr) > 12:
                     return False
@@ -542,7 +542,8 @@ class Bed(Interval):
                         return False
                     if len(block_sizes) != block_count or len(block_starts) != block_count:
                         return False
-            return True
+                found_valid_lines = True
+            return found_valid_lines
         except Exception:
             return False
 
@@ -818,28 +819,33 @@ class Gff(Tabular, _RemoteCallMixin):
         if len(get_headers(file_prefix, '\t', count=2)) < 2:
             return False
         try:
-            headers = iter_headers(file_prefix, '\t')
-            for hdr in headers:
-                if hdr and hdr[0].startswith('##gff-version') and hdr[0].find('2') < 0:
+            found_valid_lines = False
+            for hdr in iter_headers(file_prefix, '\t'):
+                if not hdr or hdr == ['']:
+                    continue
+                if hdr[0].startswith('##gff-version') and hdr[0].find('2') < 0:
                     return False
-                if hdr and hdr[0] and not hdr[0].startswith('#'):
-                    if len(hdr) != 9:
-                        return False
+                # The gff-version header comment may have been stripped, so inspect the data
+                if hdr[0].startswith('#'):
+                    continue
+                if len(hdr) != 9:
+                    return False
+                try:
+                    int(hdr[3])
+                    int(hdr[4])
+                except Exception:
+                    return False
+                if hdr[5] != '.':
                     try:
-                        int(hdr[3])
-                        int(hdr[4])
+                        float(hdr[5])
                     except Exception:
                         return False
-                    if hdr[5] != '.':
-                        try:
-                            float(hdr[5])
-                        except Exception:
-                            return False
-                    if hdr[6] not in data.valid_strand:
-                        return False
-                    if hdr[7] not in self.valid_gff_frame:
-                        return False
-            return True
+                if hdr[6] not in data.valid_strand:
+                    return False
+                if hdr[7] not in self.valid_gff_frame:
+                    return False
+                found_valid_lines = True
+            return found_valid_lines
         except Exception:
             return False
 
@@ -953,37 +959,41 @@ class Gff3(Gff):
         if len(get_headers(file_prefix, '\t', count=2)) < 2:
             return False
         try:
-            headers = iter_headers(file_prefix, '\t')
-            for hdr in headers:
-                if hdr and hdr[0].startswith('##gff-version') and hdr[0].find('3') >= 0:
+            found_valid_lines = False
+            for hdr in iter_headers(file_prefix, '\t'):
+                if not hdr or hdr == ['']:
+                    continue
+                if hdr[0].startswith('##gff-version') and hdr[0].find('3') >= 0:
                     return True
-                elif hdr and hdr[0].startswith('##gff-version') and hdr[0].find('3') < 0:
+                elif hdr[0].startswith('##gff-version') and hdr[0].find('3') < 0:
                     return False
-                # Header comments may have been stripped, so inspect the data
-                if hdr and hdr[0] and not hdr[0].startswith('#'):
-                    if len(hdr) != 9:
+                # The gff-version header comment may have been stripped, so inspect the data
+                if hdr[0].startswith('#'):
+                    continue
+                if len(hdr) != 9:
+                    return False
+                try:
+                    int(hdr[3])
+                except Exception:
+                    if hdr[3] != '.':
                         return False
+                try:
+                    int(hdr[4])
+                except Exception:
+                    if hdr[4] != '.':
+                        return False
+                if hdr[5] != '.':
                     try:
-                        int(hdr[3])
+                        float(hdr[5])
                     except Exception:
-                        if hdr[3] != '.':
-                            return False
-                    try:
-                        int(hdr[4])
-                    except Exception:
-                        if hdr[4] != '.':
-                            return False
-                    if hdr[5] != '.':
-                        try:
-                            float(hdr[5])
-                        except Exception:
-                            return False
-                    if hdr[6] not in self.valid_gff3_strand:
                         return False
-                    if hdr[7] not in self.valid_gff3_phase:
-                        return False
-                    parse_gff3_attributes(hdr[8])
-            return True
+                if hdr[6] not in self.valid_gff3_strand:
+                    return False
+                if hdr[7] not in self.valid_gff3_phase:
+                    return False
+                parse_gff3_attributes(hdr[8])
+                found_valid_lines = True
+            return found_valid_lines
         except Exception:
             return False
 
@@ -1031,34 +1041,38 @@ class Gtf(Gff):
         if len(get_headers(file_prefix, '\t', count=2)) < 2:
             return False
         try:
-            headers = iter_headers(file_prefix, '\t')
-            for hdr in headers:
-                if hdr and hdr[0].startswith('##gff-version') and hdr[0].find('2') < 0:
+            found_valid_lines = False
+            for hdr in iter_headers(file_prefix, '\t'):
+                if not hdr or hdr == ['']:
+                    continue
+                if hdr[0].startswith('##gff-version') and hdr[0].find('2') < 0:
                     return False
-                if hdr and hdr[0] and not hdr[0].startswith('#'):
-                    if len(hdr) != 9:
-                        return False
+                # The gff-version header comment may have been stripped, so inspect the data
+                if hdr[0].startswith('#'):
+                    continue
+                if len(hdr) != 9:
+                    return False
+                try:
+                    int(hdr[3])
+                    int(hdr[4])
+                except Exception:
+                    return False
+                if hdr[5] != '.':
                     try:
-                        int(hdr[3])
-                        int(hdr[4])
+                        float(hdr[5])
                     except Exception:
                         return False
-                    if hdr[5] != '.':
-                        try:
-                            float(hdr[5])
-                        except Exception:
-                            return False
-                    if hdr[6] not in data.valid_strand:
-                        return False
-                    if hdr[7] not in self.valid_gff_frame:
-                        return False
-
-                    # Check attributes for gene_id (transcript_id is also mandatory
-                    # but not for genes)
-                    attributes = parse_gff_attributes(hdr[8])
-                    if 'gene_id' not in attributes:
-                        return False
-            return True
+                if hdr[6] not in data.valid_strand:
+                    return False
+                if hdr[7] not in self.valid_gff_frame:
+                    return False
+                # Check attributes for gene_id (transcript_id is also mandatory
+                # but not for genes)
+                attributes = parse_gff_attributes(hdr[8])
+                if 'gene_id' not in attributes:
+                    return False
+                found_valid_lines = True
+            return found_valid_lines
         except Exception:
             return False
 
diff --git a/lib/galaxy/datatypes/sniff.py b/lib/galaxy/datatypes/sniff.py
index 272042ece26..a63d5340bb9 100644
--- a/lib/galaxy/datatypes/sniff.py
+++ b/lib/galaxy/datatypes/sniff.py
@@ -235,24 +235,22 @@ def is_column_based(fname_or_file_prefix, sep='\t', skip=0):
         return False
 
     try:
-        headers = get_headers(fname_or_file_prefix, sep)
+        headers = get_headers(fname_or_file_prefix, sep, comment_designator='#')[skip:]
     except UnicodeDecodeError:
         return False
     count = 0
     if not headers:
         return False
-    for hdr in headers[skip:]:
-        if hdr and hdr[0] and not hdr[0].startswith('#'):
-            if len(hdr) > 1:
+    for hdr in headers:
+        if hdr and hdr != ['']:
+            if count:
+                if len(hdr) != count:
+                    return False
+            else:
                 count = len(hdr)
-            break
-    if count < 2:
-        return False
-    for hdr in headers[skip:]:
-        if hdr and hdr[0] and not hdr[0].startswith('#'):
-            if len(hdr) != count:
-                return False
-    return True
+                if count < 2:
+                    return False
+    return count >= 2
 
 
 def guess_ext(fname, sniff_order, is_binary=False):
@@ -303,13 +301,13 @@ def guess_ext(fname, sniff_order, is_binary=False):
     >>> guess_ext(fname, sniff_order)
     'gff3'
     >>> fname = get_test_fname('2.txt')
-    >>> guess_ext(fname, sniff_order)  # 2.txt
+    >>> guess_ext(fname, sniff_order)
     'txt'
     >>> fname = get_test_fname('2.tabular')
     >>> guess_ext(fname, sniff_order)
     'tabular'
     >>> fname = get_test_fname('3.txt')
-    >>> guess_ext(fname, sniff_order)  # 3.txt
+    >>> guess_ext(fname, sniff_order)
     'txt'
     >>> fname = get_test_fname('test_tab1.tabular')
     >>> guess_ext(fname, sniff_order)
@@ -451,6 +449,9 @@ def guess_ext(fname, sniff_order, is_binary=False):
     >>> fname = get_test_fname('1imzml')
     >>> guess_ext(fname, sniff_order)  # This test case is ensuring doesn't throw exception, actual value could change if non-utf encoding handling improves.
     'data'
+    >>> fname = get_test_fname('too_many_comments_gff3.tabular')
+    >>> guess_ext(fname, sniff_order)  # It's a VCF but is sniffed as tabular because of the limit on the number of header lines we read
+    'tabular'
     """
     file_prefix = FilePrefix(fname)
     file_ext = run_sniffers_raw(file_prefix, sniff_order, is_binary)
diff --git a/lib/galaxy/datatypes/test/too_many_comments_gff3.tabular b/lib/galaxy/datatypes/test/too_many_comments_gff3.tabular
new file mode 100644
index 00000000000..e3fb5113dc6
--- /dev/null
+++ b/lib/galaxy/datatypes/test/too_many_comments_gff3.tabular
@@ -0,0 +1,65 @@
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+#
+ctgA	est	match	5410	7503	.	-	.	ID=EST:agt830.3;Target=agt830.3+1+595
+ctgA	est	HSP	7000	7503	.	-	.	Parent=EST:agt830.3;Target=agt830.3+1+504
+ctgA	est	HSP	5410	5500	.	?	.	Parent=EST:agt830.3;Target=agt830.3+505+595