I’m working on a site where these bugs randomly appear in Safari and Chrome:
Safari http://img163.imageshack.us/img163/7860/screenshot20100217at314.jpg
i.e. tags appear not to have been closed off completely. The problem is when I view the source I discover it’s completely valid.
Edit
Here’s the offending HTML which I saved: http://www.pastebin.com/f1e0edd24
Are you SURE it is completely valid? Also, whichever editor you view it in might be hiding something from you. It looks like something got escaped or encoded that shouldn’t have been. Take a look at the source for the list item that is bad and compare that to a valid one. Try and find the difference between them.
Also, view the response in Fiddler to analyze the actual response html before it hits the browser. This should help reveal if it is actually the browser or is some escaping/encoding issue like I mentioned before.