From the SeeIT.org blog today there's a new post rehashing an old topic that pops up from time to time in the PHP community - the include vs include_once performance debate.
The conventional wisdom always said that PHP’s include()/require() was quicker than include_once()/require_once(), but recently I came across an interesting post by Arin Sarkissian which suggests otherwise. [...] So in keeping with the spirit of quick and dirty experimentation I hacked up some code and ran some tests on include()/require() against include_once()/require_once() and on the relative/absolute path issue. The results are pretty surprising and I love to hear some views.
He includes the testing methodology consisted of creating 10,000 files with a simple variable assignment in them and a loop to run through all of them to record the time (as reported by microtime) switching out the different times of inclusion each time - include, include_once, require, require_once. You'll have to visit the post yourself to see the results of the tests. There's charts and tables showing the differences in results based on things like using APC and PHP4 versus PHP5.