Canonicalization solves the duplicate-content problem by declaring a preferred URL when multiple URLs serve substantially similar content. The mechanism is the `<link rel="canonical" href="...">` tag in the document head, which signals to Google that this page should be treated as a duplicate of the URL specified. Common canonicalization scenarios: (1) Parameter variants (page.com/product?color=red points to page.com/product); (2) Print-friendly versions; (3) Mobile and desktop URL variants (largely obsolete since responsive design); (4) HTTP/HTTPS and www/non-www consolidation; (5) Syndicated content where another publication legally republished your article; (6) Faceted-navigation pages that should not have separate ranking identities. Common canonicalization mistakes: (1) Self-canonicalizing every page incorrectly (not wrong but unnecessary); (2) Canonicalizing pages to their own pagination (page 2 should self-canonicalize, not point to page 1); (3) Cross-domain canonicals to the wrong destination, accidentally giving away your content's ranking signals; (4) Conflicting signals (canonical says A but internal links and sitemap say B) — Google may ignore the canonical entirely. Audit canonicals quarterly with Screaming Frog or Sitebulb.