Also, noticed that CURLOPT_HTTPHEADER and CURLOPT_HEADER have the same value.
Using curl -H "UID: spud" http://... works properly.
Except for CURLOPT_HTTPHEADER and chopping off 2 bytes at the end of a CURLOPT_RETURNTRANSFER ... everything works great.
Here's my script:
<?php
$url = "http://localhost/perl-bin/mod_perl_test.pl";
$user_agent = "VanillaZilla/2001";

$ch = curl_init();
curl_setopt ($ch, CURLOPT_HTTPHEADER, "UID: spud");
curl_setopt ($ch, 23, "UID: spud");
echo "after CURLOPT_HTTPHEADER(".CURLOPT_HTTPHEADER." CURLOPT_HEADER=".CURLOPT_HEADER."):<br>";
echo "curl_errno=". curl_errno($ch) ."<br>";
echo "curl_error=". curl_error($ch) ."<br>";
curl_setopt ($ch, CURLOPT_VERBOSE, 1);
curl_setopt ($ch, CURLOPT_URL, $url);
curl_setopt ($ch, CURLOPT_HEADER, 1);
curl_setopt ($ch, CURLOPT_USERAGENT, $user_agent);
curl_setopt ($ch, CURLOPT_COOKIE, "spud=tater;tater=spud"); 
curl_setopt ($ch, CURLOPT_RETURNTRANSFER, 1);
curl_setopt ($ch, CURLOPT_POST, 1);
curl_setopt ($ch, CURLOPT_POSTFIELDS, "pass=spud&data=stuff&case=en");
$result = curl_exec ($ch);
echo "after curl_exec:<br>";
echo "curl_errno=". curl_errno($ch) ."<br>";
echo "curl_error=". curl_error($ch) ."<br>";
curl_close ($ch);

print "<hr><pre>";
echo $result . "<br>";
print "</pre>";
?>
thanks.
chris  cleesmith@hotmail.com