I hope this perl script will solve your purpose Usage : perl script.pl Id_file file.fasta Output.fasta

#!/usr/bin/perl

@file1 = `cat $ARGV[0]`;
@file2 = `cat $ARGV[1]`;
%hash;
$id ;
foreach $fasta(@file2){
        chomp($fasta);
        if($fasta =~ '>'){
                $id = $fasta;}
        if($fasta !~ '>'){
                push @{$hash{$id}},$fasta;      }
}
open(OUT,">$ARGV[2]") or die;
foreach $id(@file1){
        chomp($id);
        $id =~s/^/>/;
        #print "$id\n";
        @seq;
        print  OUT "$id\n";
        foreach $key(keys %hash){
                if($id eq $key){
                        @seq = @{$hash{$key}};}}
                for($i =0; $i<scalar(@seq);$i++){
                        print OUT "$seq[$i]\n";}
}

Hello,

we are currently developing SEDA http://www.sing-group.org/seda, a multi-platform application for processing and filtering FASTA files. I guess that what you need can be achieved with the "Pattern filtering" option, section 3.2 of the manual: http://www.sing-group.org/seda/downloads/manuals/seda-user-manual-1.0.0.pdf. In this option, you should select "Header" and then load your ID list using the "Import patterns" button. This way only sequences that contains a ID in your list will be kept.

Regards,

Hugo.

I am certain faSomeRecords utility from Jim Kent (UCSC) will do what you need.