Oh, heh. Delete non-duplicates. Right, sorry. :)

Anyway, what you have posted works fine (which is what I thought initially, before I got confused about which you were wanting to remove).

[johnsca@CORY tmp]$ cat tst.pl
#!/usr/bin/perl

my @Tapes = (
  ['Label 1', 'p', 'l', 'P'],
  ['Label 1', 'p', 'l', 'P'], # Duplicate
  ['Label 2', 'p', 'l', 'P'],
  ['Label 3', 'p', 'l', 'P'],
);

hash_tapes(@Tapes);

sub hash_tapes (@) {

    my @tapes = @_;
    my %Tape_Cat =();

    foreach my $tape (@tapes) {
        my ( $label, $protection, $location, $pool ) = @{$tape};
        push @{ $Tape_Cat{$label} }, [ $protection, $location, $pool ]
+;
    }

    # Get rid of non-duplicates
    my $count = 0;
    foreach my $tape (@tapes) {
        my ($label,$protection,$location,$pool) = @{$tape};
        if ( @{$Tape_Cat{ $label}} < 2 ) {
            print "Deleting ". $label.
                  "Because it has ".@{$Tape_Cat{ $label }}.
                  " record(s)\n";
            $count++;
            delete $Tape_Cat{$label}
        }
    }

    my @keys = keys %Tape_Cat;
    print "Deleted $count keys, keys: @keys\n";
    return \%Tape_Cat;
}
[johnsca@CORY tmp]$ ./tst.pl
Deleting Label 2Because it has 1 record(s)
Deleting Label 3Because it has 1 record(s)
Deleted 2 keys, keys: Label 1
[download]

Correct, no? You should probably double-check the data that this function is actually getting. Use Data::Dumper at the begining of the function, or the built-in perl debugger. Update: If the data is coming from an outside source, check for strange things like leading or trailing newlines or linebreaks, or non-printable characters, or such.

Good luck. :)

bbfu
Black flowers blossum
Fearless on my breath