TL;DR: Dofollow links pass ranking authority. Nofollow links used to pass nothing, until Google changed the rules in 2019. Now nofollow is a "hint," not a directive, and Google may pass value through nofollow links at their discretion. For your link building strategy, this means: stop dismissing nofollow links from high-authority sites. For your internal linking: don't bother with nofollow on internal links (reduce unnecessary links instead). The 2024 API leak confirmed PageRank is still active, which means link attributes still matter.
I spent the first five years of my SEO career treating nofollow links as worthless. If a guest post opportunity came back with "we nofollow outbound links," I'd pass. Looking back at the data, that was wrong, and the 2019 changes made it obvious. This guide reflects what I've learned after watching link data across 12,000+ sites (audited between 2024 and early 2026) and realizing that the dofollow/nofollow binary most of us learned is an oversimplification.
| Attribute | What It Does | Who Controls It |
|---|---|---|
| Dofollow (no rel attribute) | Passes full ranking authority to the linked page | The linking site |
Nofollow (rel="nofollow") |
Hints to Google: "don't pass authority through this link" | The linking site |
That's the one-sentence version. The nuance (and there's meaningful nuance that most guides gloss over) is in how Google actually treats these attributes in 2026, which has diverged significantly from what the SEO industry still teaches.
The nofollow attribute has had four distinct eras, and advice from any era before the current one is actively misleading. Knowing which era someone's advice comes from tells you whether to trust it.
Google, Yahoo, and Microsoft jointly introduced rel="nofollow" in January 2005. The specific problem: blog comment spam. Spammers were flooding every blog comment section and forum with links to their sites, extracting PageRank from legitimate sites at scale. Nofollow gave site owners a way to say "I'm hosting this link but I don't vouch for it."
The understanding was simple and absolute: nofollow links pass zero PageRank. No crawling, no indexing signal, no value whatsoever. This was a directive, not a suggestion.
SEOs quickly realized they could use nofollow strategically on their own sites. The technique was called "PageRank sculpting": if you nofollowed links to your privacy policy, login page, and other low-value pages, the PageRank that would have flowed to those pages would redistribute to your remaining dofollow links. More juice to the pages that matter.
It worked. Until Matt Cutts announced in June 2009 that it didn't anymore. Google changed the behavior: PageRank assigned to nofollowed links now evaporates rather than redistributing. If your page has 10 links and you nofollow 5, the remaining 5 don't get extra juice. The nofollowed juice simply disappears.
This was a significant change that many SEOs missed at the time (honestly, I missed it too for longer than I'd like to admit). I still encounter sites in 2026 that nofollow their internal utility links thinking they're sculpting PageRank. They're not. They're just wasting it. I covered this in detail in our link juice guide if you want the full mechanics.
This is the big one. In September 2019, Google fundamentally changed how nofollow works:
rel="sponsored" for paid links and rel="ugc" for user-generated contentThe practical implication: nofollow links from high-authority sites probably pass some value. How much? Google hasn't said. My honest assessment is that it's meaningfully less than dofollow but meaningfully more than zero, especially from authoritative, relevant sources. Gary Illyes said at the time that they needed the flexibility to "not lose valuable link signals," which implies they were already seeing important signals being lost behind nofollow walls. John Mueller later confirmed that Google uses nofollow links for discovery and "other purposes" without elaborating.
In May 2024, a massive leak of Google's internal API documentation revealed that Google still actively computes and uses PageRank. The leaked attributes included RawPageRank, PageRank2, and PageRank_NS (Rand Fishkin and Mike King at iPullRank published the most thorough analyses of what these fields mean). While the leak didn't specifically detail how nofollow factors into these calculations, it confirmed that link-based authority signals remain central to ranking. The dofollow/nofollow distinction still carries real weight in the algorithm.
Since 2019, there are four distinct link behaviors. Most guides lump them into "dofollow" and "everything else." That simplification leads to wrong decisions, particularly around sponsored content where the wrong attribute choice can trigger a manual action.
| Attribute | HTML | Use When | Authority Transfer | Google Treatment |
|---|---|---|---|---|
| Dofollow | <a href="..."> |
You trust the page and want it to rank | Full | Directive — always followed |
| Nofollow | <a href="..." rel="nofollow"> |
Untrusted or unvetted links; general "don't vouch" | Minimal to moderate (hint) | Hint — may still pass value |
| Sponsored | <a href="..." rel="sponsored"> |
Paid placements, affiliate links, sponsorship deals | Minimal (hint) | Hint — signals paid relationship |
| UGC | <a href="..." rel="ugc"> |
Comments, forum posts, user profiles, reviews | Minimal (hint) | Hint — signals user-generated origin |
When you're adding a link to your site, here's how I'd think about it:
rel="sponsored". This is important: using dofollow on paid links violates Google's guidelines and risks manual action. Sponsored is safer than nofollow here because it's more specific about the relationship.rel="ugc". If your CMS doesn't support ugc specifically, nofollow is acceptable as a fallback.rel="nofollow". When in doubt, nofollow. The downside is minimal.Yes. rel="nofollow sponsored" is valid and sometimes appropriate for paid links where you also want to explicitly signal "don't pass value." Google treats combined attributes as the union of their individual signals. In practice, rel="sponsored" alone is sufficient for paid links. Adding nofollow is redundant but not harmful.
Should you nofollow any internal links? There are two camps, and I've changed my position on this over time. Three years ago I was nofollowing our own login and pricing pages. Now I think that was wrong, and the data convinced me.
The logic: pages like your login page, privacy policy, terms of service, and shopping cart don't need ranking authority. Nofollowing links to them signals to Google which pages you consider important. Some SEOs still advocate for this, and I understand the intuition.
Three reasons:
The better approach: audit your navigation and footer links. Remove ones that don't serve users. For the ones that need to exist for usability (login, cart), leave them dofollow but consider whether they need to appear in the global nav on every page, or whether they can live in more targeted locations.
I'll acknowledge the counterargument: some very large sites (millions of pages with complex crawl budget constraints) may benefit from targeted nofollow on internal links to manage crawling priorities. If you're at that scale, you probably have a technical SEO team making these decisions based on log file analysis. For sites under 10,000 pages, internal nofollow is almost never the right move.
This is where the 2019 change has the most practical impact on day-to-day SEO work.
I've seen agencies turn down guest post opportunities, interview features, and resource page inclusions because the link would be nofollow. Before 2019, that calculus made sense. Nofollow links passed zero value. In 2026, that's leaving value on the table.
A nofollow link from The New York Times, TechCrunch, or a major industry publication carries signals that go beyond raw PageRank:
The ROI calculus has shifted. A nofollow link from a DR 90 publication is almost certainly more valuable than a dofollow link from a DR 20 guest post farm. Yet many link builders optimize purely for dofollow, chasing low-quality sites that happen to use dofollow while ignoring high-quality opportunities behind nofollow.
Wikipedia nofollows all external links. Despite this, pages cited as Wikipedia sources consistently rank well. Correlation isn't causation (Wikipedia tends to cite authoritative sources that would rank anyway), but there's growing consensus that Wikipedia nofollow links carry unusual weight. Possibly because Google gives them special treatment as a notability signal rather than a pure link signal.
My position: getting cited in Wikipedia is worth pursuing for brand authority and potential indirect ranking benefit, regardless of the nofollow. But pursue it by being genuinely notable and citation-worthy, not by editing Wikipedia articles yourself. That path leads to bans and public embarrassment. I've watched two SaaS companies get called out on Twitter for it.
A healthy backlink profile in 2026 includes both dofollow and nofollow links. In our analysis across 12,000+ sites (audited 2024-2026), top-performing domains typically show:
A profile that's 95%+ dofollow is actually unusual for natural link acquisition (I know, counterintuitive, but think about it). It often signals that links were deliberately acquired, because most large publications, social platforms, and user-generated platforms default to nofollow. Ironically, having too many dofollow links can itself be a red flag.
In our audits, roughly 8% of sites have internal nofollow links that serve no purpose. Typically leftover from outdated SEO advice or CMS plugins (especially older WordPress SEO plugins that shipped with "nofollow category/tag pages" enabled by default). These are easy wins: remove the nofollow attribute and let the authority flow naturally.
Also worth auditing: are your outbound links properly attributed?
rel="sponsored" (or at minimum, nofollow)rel="ugc"We use rel="nofollow noopener" target="_blank" on all external links at SEOJuice. It's a blanket policy that prioritizes our own authority retention. Whether that's the right call for every site depends on your link philosophy. Some SEOs (Marie Haynes and others in the E-E-A-T camp) believe generous dofollow outbound linking builds topical trust with Google. I'm not convinced the data supports that position, but I acknowledge the argument exists and hasn't been definitively settled.
A nofollow link is a hyperlink with rel="nofollow" in its HTML tag. It signals to search engines: "I'm linking to this page but I don't want to vouch for it with my ranking authority." Since 2019, Google treats this as a hint rather than a strict directive, meaning they may still pass some value through nofollow links at their discretion. The attribute was originally created in January 2005 as a joint effort by Google, Yahoo, and Microsoft to combat blog comment spam.
More than they used to, less than dofollow. Before 2019, nofollow links passed zero ranking value by design. Since Google changed nofollow to a "hint," some value likely passes through, especially from high-authority, relevant sources. Beyond direct ranking value, nofollow links drive referral traffic, build brand awareness, and signal notability to algorithms. A nofollow link from a major publication is worth more than a dofollow link from a low-quality directory.
Dofollow links (technically just links without a rel attribute) pass full ranking authority from the linking page to the target page. Nofollow links include rel="nofollow" and signal that the linking site doesn't want to pass authority. In practice, the difference has narrowed since 2019. Nofollow may still pass some value, but dofollow remains significantly stronger for direct ranking impact.
It depends on your strategy. We do it at SEOJuice (we'd rather keep our authority internal), and many sites follow this approach. The counterargument: some SEOs believe generous outbound linking to authoritative sources builds topical trust with Google. There's no definitive public data settling this debate. If you're unsure, nofollow is the safer default. The downside (slightly less perceived trust by Google, if that's even real) is smaller than the upside of retaining authority you've earned.
Add rel="nofollow" to the anchor tag: <a href="https://example.com" rel="nofollow">link text</a>. For external links, best practice is to also add noopener (security) and target="_blank" (opens in new tab): <a href="https://example.com" rel="nofollow noopener" target="_blank">link text</a>. Most CMS platforms have settings or plugins to handle this automatically for external links.
The dofollow/nofollow distinction matters less than it did five years ago, and that's probably a good thing. Google's shift toward treating nofollow as a hint means the web's link graph is less artificially fragmented. For your strategy: pursue quality links regardless of attribute, use descriptive anchor text, and stop wasting energy trying to nofollow-sculpt your internal links. The authority you're trying to redirect just evaporates. Cut unnecessary links instead.
The question I'm still turning over: as Google gets better at understanding link intent through contextual signals, will the explicit dofollow/nofollow distinction eventually become meaningless? If Google can already infer whether a link is editorial, paid, or user-generated from context alone, maybe the attributes become redundant. We're not there yet. But the trajectory suggests the distinction will keep narrowing.
If you want to see where your internal links are leaking authority through unnecessary nofollows, run a free site audit. It flags every internal nofollow and shows you exactly how much PageRank you're evaporating.
Related reading:
no credit card required