strtoupper

One might think that setting the correct locale would do the trick with for example german umlauts, but this is not the case. You have to use mb_strtoupper() instead:
<?php
setlocale(LC_CTYPE, 'de_DE.UTF8');
echo strtoupper('Umlaute äöü in uppercase'); // outputs "UMLAUTE äöü IN UPPERCASE"
echo mb_strtoupper('Umlaute äöü in uppercase', 'UTF-8'); // outputs "UMLAUTE ÄÖÜ IN UPPERCASE"
?>

It has been mentioned in a previous comment that all you need to do to let PHP's strtoupper() do the conversion - instead of writing more or less complicated functions yourself - is to specify the locale in which you're doing the case conversion:
<?php setlocale(LC_CTYPE, "de_AT") ?>

It is important to note that setlocale() will silently fail if it can't find the specified locale on your system, so *always* check its return value. Try different spellings: using "de_AT" as an example, there are various combinations that may or may not work for you: "de", "de_AT.utf8", "de_AT.iso-8859-1", "de_AT.latin1", "de_AT@euro", etc).
If you can't find an appropriate locale setting, check your system configuration (locales are a system-wide setting, PHP gets them from the OS). On Windows, locales can be set from the Control Panel; on Linux it depends on your distribution. You can try "sudo dpkg-reconfigure locales" on Debian-based distros, or configure them manually. On Ubuntu Dapper, I had to copy entries over from /usr/share/i18n/SUPPORTED to /var/lib/locales/supported.d/local, then do the dpkg-reconfigure.
After you're done, restart the web server.
That said, there are special cases where you want to do the conversion manually. In German, for example, the letter 'ß' (szlig) only exists as a lower-case character, and so doesn't get converted by strtoupper. The convential way to express a 'ß' in an uppercase string is "SS". This function will take care of this exception (for Latin1 and most of Latin9, at least):
<?php
define("LATIN1_UC_CHARS", "ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖØÙÚÛÜÝ");
define("LATIN1_LC_CHARS", "àáâãäåæçèéêëìíîïðñòóôõöøùúûüý");
function uc_latin1 ($str) {
 $str = strtoupper(strtr($str, LATIN1_LC_CHARS, LATIN1_UC_CHARS));
 return strtr($str, array("ß" => "SS"));
}
?>

something I myself first not thought about:
if there are any html entities (named entities) in your string, strtoupper will turn all letters within this entities to upper case, too. So if you want to manipulate a string with strtoupper it should contain only unicode entities (if ever).

If you only need to extend the conversion by the characters of a certain language, it's possible to control this using an environment variable to change the locale:
setlocale(LC_CTYPE, "de_DE");

In the Irish language certain initial mutations can never be capitalized — the following simple function can be used to capitalize text in Irish.
i.e. Muintir na hÉireann -> MUINTIR NA hÉIREANN
<?php
function strtoupper_ga($a) {
 return strtr(mb_strtoupper($a, "utf-8"), array(
 " MB" => " mB", 
 " GC" => " gC", 
 " ND" => " nD", 
 " BHF" => " bhF", 
 " NG" => " nG", 
 " BP" => " bP", 
 " DT" => " dT", 
 " HA" => " hA",
 " HE" => " hE",
 " HI" => " hI",
 " HO" => " hO",
 " HU" => " hU",
 " HÁ" => " hÁ",
 " HÉ" => " hÉ",
 " HÍ" => " hÍ",
 " HÓ" => " hÓ",
 " HÚ" => " hÚ"
 ));
}
?>

Here is how to make the character in upper case, except HTML-entities:
$s = substr(preg_replace('/(?<=^|;)(.+?)(?=&[0-9A-Za-z]+;|$)/e', "strtoupper('1ドル')", ' '.$s), 1);
There is small kludge, however. Unfortunately I tired to find out the way how to exclude HTML-entity at the start of the line, so I have added 1 dummy character at the start of the text and removing it after the conversion.

// 2005年5月30日 Justin
 // Chinese_Traditional toupper
 function CT_to_upper($string)
 { 
 $isChineseStart = false;
 
 $new_string = "";
 $i = 0;
 while($i < strlen($string))
 { 
 if (ord(substr($string,$i,1)) <128)
 {
 if( $isChineseStart == false )
 $new_string .= strtoupper(mb_substr($string,$i,1));
 else 
 $new_string .= substr($string,$i,1);
 }
 else
 {
 if( $isChineseStart == false )
 $isChineseStart = true;
 else
 $isChineseStart = false; 
 
 $new_string .= substr($string,$i,1);
 }
 $i++;
 }
 return $new_string; 
 } 
 //

perfect solutions for turkish utf-8 (including i I conversations):
<?php
function strtolowertr($metin){
 return mb_convert_case(str_replace('I','ı',$metin), MB_CASE_LOWER, "UTF-8");
}
function strtouppertr($metin){
 return mb_convert_case(str_replace('i','İ',$metin), MB_CASE_UPPER, "UTF-8");
}
function ucwordstr($metin) {
 return ltrim(mb_convert_case(str_replace(array(' I',' ı', ' İ', ' i'),array(' I',' I',' İ',' İ'),' '.$metin), MB_CASE_TITLE, "UTF-8"));
}
function ucfirsttr($metin) {
 $metin = in_array(crc32($metin[0]),array(1309403428, -797999993, 957143474)) ? array(strtouppertr(substr($metin,0,2)),substr($metin,2)) : array(strtouppertr($metin[0]),substr($metin,1));
return $metin[0].$metin[1];
}
?>

When using UTF-8 and need to convert to uppercase with 
special characters like the german ä,ö,ü (didn't test for french,polish,russian but think it should work, too) try this:
function strtoupper_utf8($string){
 $string=utf8_decode($string);
 $string=strtoupper($string);
 $string=utf8_encode($string);
 return $string;
}

Simple function to change the case of your string and any accented html characters contained within it. 
Inspired by fullUpper(), by silent at gmx dot li... just a little bit more atomic.
<?php
function convertCase($str, $case = 'upper')
{ //yours, courtesy of table4.com :)
 switch($case)
 {
 case "upper" :
 default:
 $str = strtoupper($str);
 $pattern = '/&([A-Z])(UML|ACUTE|CIRC|TILDE|RING|';
 $pattern .= 'ELIG|GRAVE|SLASH|HORN|CEDIL|TH);/e';
 $replace = "'&'.'\1円'.strtolower('\2円').';'"; //convert the important bit back to lower
 break;
 
 case "lower" :
 $str = strtolower($str);
 break;
 }
 
 $str = preg_replace($pattern, $replace, $str);
 return $str;
}
?>

Depending on what you are trying to achieve you would call like this:
<?php
//with entities...
$str = convertCase(htmlentities($str, ENT_QUOTES, "ISO-8859-1"));
?>

Russian
function str_to_upper($str){
 return strtr($str, 
 "abcdefghijklmnopqrstuvwxyz".
 "\xE0\xE1\xE2\xE3\xE4\xE5".
 "\xb8\xe6\xe7\xe8\xe9\xea".
 "\xeb\xeC\xeD\xeE\xeF\xf0".
 "\xf1\xf2\xf3\xf4\xf5\xf6".
 "\xf7\xf8\xf9\xfA\xfB\xfC".
 "\xfD\xfE\xfF",
 "ABCDEFGHIJKLMNOPQRSTUVWXYZ".
 "\xC0\xC1\xC2\xC3\xC4\xC5".
 "\xA8\xC6\xC7\xC8\xC9\xCA".
 "\xCB\xCC\xCD\xCE\xCF\xD0".
 "\xD1\xD2\xD3\xD4\xD5\xD6".
 "\xD7\xD8\xD9\xDA\xDB\xDC".
 "\xDD\xDE\xDF");
}

$ther=''.THEREISALREADYA.' '.CONCEPT.' '.SAVED.' ';
or
$ther="There Is all ready A concept SAVED";
$fupper=substr("$ther",0,1);
pick the first char
$theru = strtoupper($fupper);
make it upper 
$flower=substr("$ther",1,100);
pick the rest
$therl = strtolower($flower);
make them lower
Result:
There is all ready a concept saved

This function has a real challenge when it comes to Turkish. In Turkish the ASCII letter 'i' uppercases to a non-ASCII character. This means that PHP cannot upper case it.
i.e. for the Turkish locale, strtoupper('i')=='i'
This can mess with basic program logic.
There's no simple solution. The core problem is discussed more here:
http://www.i18nguy.com/unicode/turkish-i18n.html 

chinese
function to_upper($string) {
 $new_string = "";
 $i = 0;
 while($i < strlen($string)) {
 if (ord(substr($string,$i,1)) <128)
 {
 $new_string .= strtoupper(substr($string,$i,1));
 $i++;
 } else {
 $new_string .= substr($string,$i,2);
 $i=$i+2;
 }
 }
 return $new_string;
}