If Google's index contains multiple URLs for my homepage, does that mean the canonical tag is not working?

[JMagary]

I have a site which is using canonical tags on all pages, however not all duplicate versions of the homepage are 301'd due to a limitation in the hosting platform.  So some site visitors get www.example.com/default.aspx while others just get www.example.com.  I can see the correct canonical tag on the source code of both versions of this homepage, but when I search Google for the specific URL "www.example.com/default.aspx" I see that they've indexed that specific URL as well as the "clean" one.  Is this a concern... shouldn't Google only show me the clean URL?

[Dr-Pete]

In most cases, Google does seem to "de-index" the non-canonical URL, if they process they tag. I put in quotes just because, technically, the page is still in Google's index - as soon as it's not showing up at all (including with "site:"), though, I essentially consider that to be de-indexed. If we can't see it, it might as well not be there.

If 301-ing isn't an option, I'd double-check a few things:

(1) Is the non-canonical page ranking for anything (including very long-tail terms)?

(2) Are there any internal links to the non-canonical URL? These can send a strongly mixed signal.

(3) Are there any other mixed signals that might be throwing off the canonical? Examples include canonicals on other pages that contradict this one, 301s/302s that override the canonical, etc.

[MikeRoberts]

As Digital-Diameter said, the best choice for fixing this problem is a 301. A Canonical tag can eventually lead to the incorrect URL being replaced by the correct one in the SERPs but it is also important to note that the Rel=canonical tag is a suggestion, not a directive. What this means is that the search engines will take it into consideration but may choose not to follow it.

[Digital-Diameter]

Technically, rel=canonical tags can still leave a page indexed, they simply pass authority for Google. From your question I can tell you know this, but I do have to say that 301's are the best way to address this. Blocking a page with robots.txt can help as well, but this just stops Google from crawling a page, the page can still be indexed again.

If you have pages or versions of pages that you do not want indexed you may want to use the no index meta tag. Google's notes here. Be careful though, this will stop these pages from being indexed, but they will still be crawled (though your rel=canonical solution should make this a non-issue).

A few other notes:

In all cases, be sure your internal links point consistently to the URL version you have determined for your home page.

WMT also creates a list of inbound links that are missing or broken. You can use this to help determine any additional 301s that you need.