isset !empty array_key_exists 厉害了我的哥

The other responses focus on the differences between the two functions. This is true, but if the source array does not contain null or 0 or "", ... (empty values) values you can benchmark the speed of the two functions:

<?php

function makeRandomArray( $length ) {
    $array = array();
    for ($i = 0; $i < $length; $i++) {
        $array[$i] = rand(1, $length);
    }

    return $array;
}

function benchmark( $count, $function ) {
    $start = microtime(true);
    for ($i = 0; $i < $count; $i++) {
        $function();
    }
    return microtime(true) - $start;
}

$runs = 100000;
$smallLength = 10;
$small = makeRandomArray($smallLength);

var_dump(benchmark($runs, function() {
    global $small, $smallLength;
    array_key_exists(rand(0, $smallLength), $small);
}));
var_dump(benchmark($runs, function() {
    global $small, $smallLength;
    !empty($small[rand(0, $smallLength)]);
}));

Which gave me the following results:

For a small array:

• array_key_exists: float(0.18357992172241)
• empty: float(0.072798013687134)
• isset: float(0.070242881774902)

For a relative big array:

• array_key_exists: float(0.57489585876465)
• empty: float(0.0068421363830566)
• isset: float(0.0069410800933838)

So if it's possible it's faster to use empty or isset.
-----------
https://stackoverflow.com/questions/6884609/array-key-existskey-array-vs-emptyarraykey