I should note, I'm not actually entirely sure this is true in every browser.  In fact, it's been so long since I even used meta refreshes that this may not be true in any browsers anymore...

But obviously the browser has to parse the entire entity body and HTML (and usually render it) before redirecting.  This is because with a refresh (unlike a redirect) the entity body is understood to be a message about why they are being redirected.

The most correct place to use a refresh these days is in front of a search results page which is slow.  The refresh would be put on a page that says "your results are being loaded."  Even so, that's unpopular these days.

-[Unknown]


Unknown W. Brackets wrote:
> 3. Well, Refresh uses a parameter, like this: "0; http://www.google.com/".  The "0" in this case is the number of seconds after the entire page has loaded (in most browsers, after the load event has been posted.)  On the other hand, a 301/Location combination happens as soon as the browser finishes reading the headers.