<b>"Use a canonical tag to keep duplicate pages out of the index."</b>

<b>"Use a canonical tag to keep duplicate pages out of the index."</b>

Wrong tool.

Canonical is a hint, not a directive — Google treats rel=canonical as a suggestion and openly ignores it when its own signals disagree (internal links, sitemap inclusion, content that's actually different). People wire up a canonical, see the page still indexed, and assume the tag is broken. It's working exactly as specified: as advice Google may decline.

Here's what actually happens: if you need a URL gone from the index, noindex is the directive that removes it. Canonical is for consolidating signals across genuine near-duplicates you're fine either way about.

The trap is mixing them: noindex plus canonical on the same page sends a contradiction ("merge into the canonical" + "drop me entirely"), and Google has warned this is conflicting.

Pick the verb you mean. Suggest, or command. Not both.