my $l;        # all these three variables should probably better decla
+red within the 
my @vals;     # while loop. Only %pairs probably need to be declared b
+efore the while
my $json;

while (<>) {
    $l = $_;
    chomp $l;
    @vals = split /;/, $l;
    if ($vals[0] =~ /Query/) {
        $pairs{$vals[1]}{$vals[2]} = $vals[3];    # %pairs isn't decla
+red anywhere
    } elsif {$vals[0] =~ /Answer/) {              # syntax error: elsi
+f { should be elsif (
        $pairs{$vals[1}{$vals[2]} = $vals[3];
        $json = encode_json $pairs{$vals[1]};     # what do you think 
+is the content of $pairs{$vals[1]}? Probably not what you want to enc
+ode.
        print $json."\n";
        delete $pairs{$vals[1]};
    }
}
[download]

Do yourself a favor. Use the following pragmas:

use strict;
use warnings;
[download]

specifically there can be anywhere from 1 to n answers for each query

See my last post at the end. I check for a repeat and when I detect it, I delete it and start over.

Update:

Hmmm, thinking about what you said and what I just said ...

Maybe deleting the entry and starting over (though needed to get rid of the last answers) does not really solve my memory issue at all. You might actually be onto something there but I do not know how to fix it, given that it is the case ...

In fact, all you need to store in your hash is the queries, there is no reason to store any answer in the hash, since you can process it immediately and print it. This may help, but I'm not sure, though, that this is sufficient to reduce your memory footprint to an acceptable level (it depends on your data profile).