##############

###retrieve member list
opendir(MEMBERS, "$free_path");
@memberz = grep !/\.\.?/, readdir MEMBERS;
closedir (MEMBERS);
&memhead;
########

    @members = sort @memberz;

if ($start eq "") { $start = 1; }

for $members (@members) {
if ($members eq "total.dat") { next; }

   $c++; # count all items
    if($c >= $start){ # if $c count is = or greater than where to star
+t
        $d++; # count items that match first test
        unless($d > $show){ # unless $d count is greater than how many
+ to show
#$action = $INPUT{'action'};
#$membername = $INPUT{'name'};
#$username = $INPUT{'username'};
#$cityx = $INPUT{'city'};
#$genderx = $INPUT{'gender'};
#$countryx = $INPUT{'country'};
#$statex = $INPUT{'state'};
#$agex = $INPUT{'age'};
#$start = $INPUT{'start'};
#$show = $INPUT{'show'};
#coollect info
use DB_File;                     # optional; overrides default
tie %cata, "DB_File", "$path/accounts" or die "Can't open FILENAME: $!
+\n";   

my $acc_data = $cata{$members};
untie(%cata);

($email,$name,$pwd,$size,$creation,$last_log,$size_added,$hf_on,
    $website_name,$website_des,$numdirs,$www_of,$www_name,$www_bgcolor
+,
    $www_bgimage,$www_text,$www_link,$www_vlink,$onhold,$address,$city
+,
    $state,$zip,$country,$telephone,$gender,$age,$icq,$education,
    $income,$occupation,$d_o_b,$moderator,$gbook_of,$op_in,$icon_url,
    $file_types,$ftp_per,$session_str,$num_dirs,$num_files,
    $custom_1,$custom_2,$custom_3,$custom_4,$custom_5,$custom_6,$custo
+m_7,
    $custom_8,$custom_9,$custom_10,$marital,$signup_ip,$last_ip,$formm
+ail_of) = split(/\|/,$acc_data);

($month,$day,$year1,$year2) = split (/\,/,$d_o_b);
#end collect info
[download]

Yes, you can still use the method suggested by dws (or my variant thereof) but you'll have to restructure your user hash like my  %users:

use strict;
my %users = ();
my %matches = ();
my %search = (
       sex => 'F',
       age => 19,
       st  => 'FL'
);

## load users hash
foreach (<DATA>) {
   chomp;
   my ($user, $sex, $age, $st) = split /\|/;
   $users{$user}{sex} = $sex;
   $users{$user}{age} = $age;
   $users{$user}{st}  = $st;
}

## Perform searches
foreach my $user (keys %users) {
   my $match = 1;
   foreach my $key (keys %search) {
      $match = 0, last if $users{$user}{$key} ne $search{$key};
   }
   $matches{$user}++ if $match;
}

## Print out the matches (if any)
foreach (keys %matches) {
   print "user: '", $_,
         "'  sex: '", $users{$_}{sex},
         "'  age: '", $users{$_}{age},
         "'  st: '",  $users{$_}{st}, "'\n";
}

print "No matches found!\n" if not %matches;

__DATA__
JOE|M|18|AL
SUE|F|18|MS
BOB|M|19|CA
EVE|F|20|FL
PAM|F|19|FL
ABE|M|18|FL
NAN|F|19|CA
[download]

Note that this could be very slow if you have a large number of users and/or a large number of criteria to match. Worst case is criteria count * user count iterations. I'm not the greatest at optimizing code, so no doubt there are more efficient ways to do it (TAMEWTDI).

--Jim