A few years ago, MaxCDN published a post on Better Image SEO with Canonical Headers in which they suggest specifying a canonical link in the HTTP header. Google's documentation on this, however, states "Google currently supports these link header elements for Web Search only." This seems to imply that it would not benefit image searches. It seems that it would be a good idea -- and the kind of thing Google would do, so I am wondering if there is a benefit to setting a canonical link for duplicate image content. Has anyone done a test of this?

Just to provide a little more specifics for my scenario, I am using a CMS that utilizes ImageProcessor.Web which can resize images and for JPEGs, reduce the size (by reducing the quality). I think it would be useful if the site served smaller, reduced quality images but had the canonical link set to the original.

  • The canonical link is an HTML construct. How are you imagining that this would work for images? Is there a proposal for this? Are there any posts you can refer to? Otherwise, are you throwing out something for discussion? Or are you looking for a construct that you are not aware of yet?
    – closetnoc
    Commented Feb 8, 2017 at 4:17
  • If the reduced/optimized image have link(like <a href="http://exa.com/high-quality-image.jpg"><img src="http://exa.com/optimize-image.jpg/></a>), and that link link point to high quality img, then G will index the high quality images. It doesn't consider as duplicate content. I don't have your answer about canonical link for images, but G does not consider that kind of images as duplicate, which is your main point. Also G know that, same images are used by many peoples(So no penalty on licence images), and they give benefits to only, which is index first, high quality, and has credit links
    – Goyllo
    Commented Feb 8, 2017 at 6:30
  • 1
    @closetnoc It's not just an HTML construct. See the Google documentation I linked to (support.google.com/webmasters/answer/…)
    – Alex
    Commented Feb 8, 2017 at 15:53
  • @Goyllo the images are embedded, not linked.
    – Alex
    Commented Feb 8, 2017 at 15:54
  • 1
    I can't confirm because PDF and other contextual content is handle by Googlebot, while for images, Googlebot-image do it's job. Both are different bot, and configured differently. That may be reason behind the statement "Google currently supports these link header elements for Web Search only. "
    – Goyllo
    Commented Feb 9, 2017 at 9:10

1 Answer 1


Short answer NO. The purpose of the canonical tag is duplicated content, not media. Google's John Mueller talks about image canonical in this office hours hangout video from Aug 12, 2016.

(56:24) For images, we don't use the rel canonical. So if you have the rel canonical header in the images themselves, we don't use that.

  • Nice find. I didn't know that. Commented Oct 23, 2018 at 15:51
  • 1
    In 2019 ther's support for rel canonical http header from Google ( support.google.com/webmasters/answer/… ) they do not explicitly mention images only PDFs but it's probabilly same with images.
    – iiic
    Commented Sep 9, 2019 at 7:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.