diff --git a/lib/PPI/Token/Pod.pm b/lib/PPI/Token/Pod.pm
index 5aac876d..6e21a326 100644
--- a/lib/PPI/Token/Pod.pm
+++ b/lib/PPI/Token/Pod.pm
@@ -132,7 +132,7 @@ sub __TOKENIZER__on_line_start {
 	# Check the line to see if it is a =cut line
 	if ( $t->{line} =~ /^=(\w+)/ ) {
 		# End of the token
-		$t->_finalize_token if lc $1 eq 'cut';
+		$t->_finalize_token if $1 eq 'cut';
 	}
 
 	0;
diff --git a/t/ppi_token_pod.t b/t/ppi_token_pod.t
index fc42cc36..ff1e225f 100644
--- a/t/ppi_token_pod.t
+++ b/t/ppi_token_pod.t
@@ -12,7 +12,7 @@ BEGIN {
 }
 use PPI;
 
-use Test::More tests => 4;
+use Test::More tests => 8;
 
 {
 	# Create the test fragments
@@ -28,4 +28,18 @@ use Test::More tests => 4;
 }
 
 
+{
+	foreach my $test (
+		[ "=pod\n=cut", [ 'PPI::Token::Pod' ] ],
+		[ "=pod\n=cut\n", [ 'PPI::Token::Pod' ] ],
+		[ "=pod\n=cut\n\n", [ 'PPI::Token::Pod', 'PPI::Token::Whitespace' ] ],
+		[ "=pod\n=Cut\n\n", [ 'PPI::Token::Pod' ] ],  # pod doesn't end, so no whitespace token
+	) {
+		my $T = PPI::Tokenizer->new( \$test->[0] );
+		my @tokens = map { ref $_ } @{ $T->all_tokens };
+		is_deeply( \@tokens, $test->[1], 'all tokens as expected' );
+	}
+}
+
+
 1;