This free checker answers “when was website last updated” for any public page: paste a URL and it collects every date the page exposes, from the HTTP Last-Modified header to JSON-LD dateModified, meta tags, visible dates, the sitemap lastmod and Wayback Machine captures. It then picks a best-guess date and tells you how confident that guess is. No method is certain on its own, so the checker shows each signal side by side and explains what each one actually measures.
The rest of this page explains those signals in depth: how to check when a website was last updated by hand, which dates you can trust, how to find when a page was published, how to cite undated pages, and how to decide whether an article is too old to rely on.
“When was website last updated?” The short answer
You find out when a website was last updated by combining several clues, because no web page is required to tell you. The most reliable clue is a date the publisher states on purpose, such as a visible “Updated” line or a dateModified value in structured data. Server headers, sitemaps and archive captures fill the gaps when the publisher says nothing.
The web has no universal “last edited” stamp. HTML has no mandatory date field, HTTP servers can omit or fake the Last-Modified header, and content management systems store edit times in their own databases where visitors can’t see them. What you can see is whatever the site chooses to expose. That’s why the question “when was website last updated” has no one-click, guaranteed answer, and why a good checker shows its working instead of a single number.
Here are the signals, ranked roughly from most to least trustworthy for the question “when did this page’s content last change?”:
| Signal | Where it lives | What it tells you | Main weakness |
|---|---|---|---|
JSON-LD dateModified |
<script type="application/ld+json"> in the page source |
The CMS’s own record of the last edit | Some themes update it on every save, even trivial ones |
| Meta modified tags | article:modified_time, og:updated_time in <head> |
Same as above, set by SEO plugins | Can be missing or stale if the plugin is misconfigured |
| Visible date | Byline or “Updated on” line near the headline | What the publisher wants readers to see | Easy to change without changing content |
Sitemap <lastmod> |
/sitemap.xml (listed in robots.txt) |
When the site says the URL last changed | Often auto-generated; may change when comments or sidebars change |
HTTP Last-Modified |
Response header from the server | When the file or cached page was last written | On dynamic pages, often just the time of the request |
| Wayback captures | web.archive.org | That the page existed, in that form, on those dates | Captures are irregular; a capture is not an edit |
document.lastModified |
Browser JavaScript property | Mirrors Last-Modified if present | Shows the current time when the header is missing |
How to check when a website was last updated (every free method)
The fastest way is to paste the URL into the tool above; the manual methods below are what it automates. Use them when you want to verify a result yourself or when a site blocks automated checks.
1. Look for a visible date on the page
Start with the page itself. Scan the area around the headline and byline for phrases like “Updated”, “Last reviewed”, “Revised” or “Modified”. Also check the bottom of the article, since documentation sites and government pages often put “Page last updated” in the footer of the content. Ignore the copyright year in the site footer: many sites update it automatically every January, so “© 2026” says nothing about the article.
A visible date is what a publisher wants you to see, which makes it both the most useful and the easiest to manipulate. Some sites refresh the displayed date whenever they touch a page, even if only a link changed.
2. Read the page source for structured dates
Press Ctrl+U on Windows, Linux and ChromeOS or Cmd+Option+U on a Mac to view the source in Chrome, Edge or Brave (Firefox uses the same shortcuts). Then press Ctrl+F or Cmd+F and search for:
dateModifiedanddatePublished, the schema.org properties inside JSON-LD blocksarticle:modified_timeandarticle:published_time, the Open Graph article tagsog:updated_time, used by some WordPress SEO plugins<time datetime=, the HTML element many themes use for the byline date
Dates here usually follow ISO 8601, such as 2026-08-14T09:30:00+00:00. The +00:00 part is the time zone offset, so a date late at night in one zone may appear as the next day in another.
3. Read the HTTP Last-Modified header
The Last-Modified header is part of HTTP itself, defined in RFC 9110. In Chrome or Edge, open DevTools (F12, or Cmd+Option+I on a Mac), switch to the Network tab, reload the page, click the first request (the document), and look under Response Headers for last-modified. In a terminal, curl -sI "https://example.com/page" | grep -i last-modified prints it directly.
A value like Last-Modified: Tue, 04 Aug 2026 11:02:17 GMT is always in GMT. If the header is missing, or it matches the exact minute you requested the page, the site generates pages on the fly and the header tells you nothing about content edits.
4. Check the sitemap lastmod
Most sites publish an XML sitemap listing their URLs, often with a <lastmod> date per URL. Open https://example.com/robots.txt and look for a Sitemap: line, or try /sitemap.xml and /sitemap_index.xml directly. Large sites split sitemaps into many files, so you may need to open the index, pick the right child sitemap (posts, pages, products), and search for your URL.
Google’s sitemap documentation says it uses <lastmod> only if it is “consistently and verifiably accurate”, and that it ignores <changefreq> and <priority>. That’s a useful hint for you too: a sitemap where every URL shows today’s date is broken, not fresh.
5. Use the Wayback Machine calendar and Changes view
The Internet Archive’s Wayback Machine has crawled the public web since 1996. Open https://web.archive.org/web/*/ followed by the URL to see a calendar of every capture. Blue and green circles mark days with snapshots; click one to see the page as it looked then.
For the question of what changed, the Wayback Machine’s Changes view at https://web.archive.org/web/changes/ plus the URL lets you pick two captures and compare them side by side, with additions and deletions highlighted. That’s the only method here that shows real content changes rather than a claimed date. The Internet Archive documents it in its developer tutorial on comparing snapshots. For more on how archives work, see our guide to website archives and how to use them, and if the Wayback Machine has no copy, archive.today sometimes does.
6. Search Google with before: and after:
Google Search supports before: and after: operators, documented on the Refine Google searches help page and still marked beta as of September 2026. A query like site:example.com "exact phrase from the page" before:2024-01-01 tells you whether Google had indexed that text before a given date. It’s a rough tool, since Google’s dates are its own estimates, but it helps when every other signal is missing.
Google removed cached pages and the cache: operator in 2024, so you can no longer open Google’s stored copy of a page to compare it with the live version.
7. The document.lastModified trick (and why it misleads)
The widely shared trick is to type javascript:alert(document.lastModified) into the address bar while the page is open. Browsers strip the javascript: prefix when you paste, so type that part by hand or run document.lastModified in the DevTools console instead.
Per MDN, the property returns the document’s last modification date and time. In practice, when the server sends no Last-Modified header, browsers return the current date and time. On almost every CMS-driven site that means the alert simply shows the moment you loaded the page. Firefox’s Page Info window (Ctrl+I, or Cmd+I on a Mac, then the General tab) shows the same value under “Modified”, with the same caveat. Treat this method as useful only for static HTML files.
Webpage last modified date: method comparison
Each method trades speed for reliability. This table sums up when to reach for which, so you can check the last modified date of a web page with the fewest steps.
| Method | Speed | Needs technical skill | Shows real edits | Best for |
|---|---|---|---|---|
| This free checker | About 10–20 seconds | No | Partly (combines signals) | A quick, sourced best guess |
| Visible date | Instant | No | Only if honest | News, blogs, documentation |
| View source (JSON-LD, meta) | 1 minute | A little | Usually | WordPress and CMS sites |
| DevTools or curl header | 1 minute | Some | Static files only | PDFs, images, plain HTML |
| Sitemap lastmod | 2–5 minutes | Some | Sometimes | Sites with well-kept sitemaps |
| Wayback calendar and Changes | 2–10 minutes | No | Yes, between captures | Verifying claims, research |
| Google before:/after: | 1 minute | No | Roughly | Pages with no other dates |
| document.lastModified | Instant | No | Rarely | Static HTML only |
Why no method can tell you the exact date for certain
No method is certain because every date signal is either written by the publisher, who can change it freely, or observed from outside at irregular intervals. The web has no neutral clock recording edits.
A few common situations show how dates go wrong:
- Cosmetic edits refresh the date. Fixing a typo or swapping an ad can bump
dateModified, so a 2019 article can look like it was “updated” last week. - Real edits leave dates untouched. Some CMS setups only change the visible date when an editor ticks a box, so substantial rewrites can hide behind an old date.
- Caching layers rewrite headers. A CDN may set Last-Modified to when it last refreshed its cache, which has nothing to do with the author.
- Templates change around content. When a site redesigns, every page’s HTML changes, the server headers and sitemap update, and the Wayback Machine records a “change”, yet the article text is identical.
- Archive captures are sparse. A popular homepage might be captured many times a day; a small blog post might be captured once a year or never. A capture only tells you the page looked a certain way on that day.
That’s why the checker reports a confidence level instead of pretending to know. High confidence means the publisher states a modified date in structured data or meta tags. Medium means the date comes from a sitemap, server header or publish date. Low means only Wayback captures were found, which give you a window (“changed no later than” or “existed by”), not an edit time.
When was a page published? Finding the original date
To find when a page was published, look for datePublished in JSON-LD, the article:published_time meta tag, or the byline date; then use the first Wayback Machine capture as a lower bound. The publish date tells you how old a page is, while the modified date tells you how fresh it is. You usually want both.
The first Wayback capture is often misread. It shows the earliest date the Internet Archive saw the page, and the page existed at least that long ago. But the Archive may have found the page months or years after it went live, especially for small sites. If a page’s datePublished is earlier than its first capture, that’s normal. If datePublished is later than the first capture, the publisher likely republished or re-dated an older page.
Other clues to the original publish date:
- The URL. Many blogs and news sites put the date in the path, such as
/2021/03/article-name/. That date usually survives later edits. - Comments. The oldest comment’s timestamp gives an upper bound for when the page went live.
- Social shares and backlinks. Posts linking to the page can’t be older than the page itself.
How to check whether an article is stale
A page is stale when its claims no longer match reality, which depends on the topic more than on its age. A 2012 explanation of how DNS works may still be accurate; a 2025 guide to a software menu may already be wrong.
Work through these checks before you rely on an article:
- Find the modified date with the checker, and compare it with the publish date. A big gap with a genuine update note is a good sign.
- Check version numbers and menu paths. If a how-to mentions a browser, app or operating system version, compare it with the current one. Software instructions go stale fastest.
- Check statistics and prices. Look for the year a figure refers to. “Costs $10 a month” without a date is a warning sign.
- Click the outbound links. Articles that cite dead sources are usually unmaintained. If you keep many research links, the free Bookmark Dead Link Checker tests a whole bookmarks export at once.
- Compare Wayback captures. Use the Changes view to confirm an “updated” date reflects real edits, not a refreshed timestamp.
- Look at the rest of the site. If the newest posts on the blog are years old, the article you’re reading probably hasn’t been touched either.
The checker adds a rough freshness badge to every result: under a year old, one to three years, or over three years. Use it as a prompt to look closer, not as a verdict.
Citing web pages: which date to use
For citations, use the date the page itself states, whether that’s the publication date or a clearly labelled update date, and fall back to “n.d.” plus a retrieval or access date when the page shows none.
| Style | Page shows a date | Page shows no date | Retrieval or access date |
|---|---|---|---|
| APA 7 | Use the most specific date shown (year, month, day); use the updated date if clearly marked | Write (n.d.) | Only for pages designed to change over time and not archived: “Retrieved September 26, 2026, from URL” |
| MLA 9 | Include the date shown | Omit the publication date | Optional, but recommended when no date is shown: “Accessed 26 Sept. 2026.” |
| Chicago (notes and bibliography) | Include the publication or revision date | Use n.d. | Include an access date when no date of publication or revision is shown |
- Don’t cite a sitemap or header date as the publication date. Those are technical signals, not dates the author put forward. Use them to decide whether to trust the page, not in the reference itself.
- Archive what you cite. Save the page to the Wayback Machine or archive.today and keep the snapshot link, so readers can see the version you read even if the page changes. The free Bulk Wayback Machine Checker shows which of your sources already have snapshots.
- Record the checker result in your notes. The “Copy summary” button under the result gives you the page URL, the best-guess date, every signal and the date you checked, in plain text.
The American Psychological Association explains its date rules in its guidance on webpage references.
Other free ways to check a page’s last updated date
Besides this page last updated checker, you can use browser developer tools, the Wayback Machine directly, Google’s date operators, browser extensions, or command-line tools.
- Browser developer tools (Chrome, Edge, Brave, Firefox, Safari). Free, built in and precise for HTTP headers and page source. Better than this checker when you need to see exactly what your own browser receives, including pages behind a login. In Safari, first turn on developer features under Safari > Settings > Advanced > “Show features for web developers” (as of September 2026).
- The Wayback Machine. Better than any checker for proving what a page said on a given day and what changed between captures. Slower, and it only covers pages it has crawled.
- Google Search operators. Useful when a page has no dates at all, but Google’s dates are estimates and the operators are officially beta.
- Browser extensions. Several free extensions show a Last-Modified value in the toolbar. They inherit the same weakness as
document.lastModifiedon dynamic sites, and each extension you install can read the pages you visit, so check permissions first. - curl or wget. Precise and scriptable for checking headers on many URLs, and good for static files. They don’t parse JSON-LD, meta tags or sitemaps unless you write that yourself.
Where this checker does better: it gathers six kinds of signals in one pass, discards the Last-Modified value when it just equals the time of the request, looks up the sitemap for you, ranks the signals, and states its confidence. It can’t see pages behind logins or paywalls, and it can’t prove a page wasn’t edited between captures.
Troubleshooting: common problems and fixes
Most problems come from the site blocking automated requests, generating dates on the fly, or not publishing dates at all.
| Symptom | Likely cause | What to do |
|---|---|---|
| The date shown is today’s date | The server generates pages on request, or a plugin refreshes dates on every load | Check the signals table: if only HTTP Last-Modified shows today, it was discarded; look at JSON-LD and Wayback rows instead |
| “No date signals found” | The page publishes no dates and has no Wayback captures | Try the URL path for a date, the oldest comment, and Google’s before: operator |
| Error saying the site couldn’t be reached | The site is down, slow, or blocks automated requests (403, Cloudflare challenge) | Open the page in your browser and use view source and DevTools manually |
| “Too many checks” message | Rate limit to keep the service free | Wait a minute and check again |
| Signals disagree by years | A redesign, a migration, or an automatic date refresh | Trust the visible content first; compare two Wayback captures with the Changes view |
| The best guess is a Wayback date with low confidence | Nothing else was found | Read it as “the page existed in this form by then”, not as an edit date |
| Checked the homepage, got a recent date | Homepages change constantly (new posts, ads, feeds) | Check the specific article URL instead |
| Short link or redirect gave a confusing result | The checker checked the final page after redirects | Expand the link first with the free URL Expander to see where it goes |
| The date is a day off | Time zone conversion from ISO 8601 or GMT | The checker shows dates in UTC; the page may display its local date |
Webpage dates and your saved bookmarks
Checking one page’s date is often part of a bigger job: deciding which of your saved links are still worth keeping. Old bookmarks don’t just go stale; they disappear. Pages get deleted, domains expire and get parked, and sites restructure their URLs, which is known as link rot. You can see the scale on your own bookmarks with the free rot check.
Bookmend doesn’t estimate update dates, but it handles the part that comes after. It’s an AI-native bookmark manager for Chrome and the web: when you save a page, it keeps the page’s readable text and writes a short summary, so the content stays available even if the site changes later. Its link health feature checks saved links in the background, at most once a day per bookmark, flags the ones that have died, and offers the Wayback Machine snapshot closest to when the page last worked as a replacement. You can also ask your library questions in plain language and get answers with citations back to your saved pages. AI features and link health are part of Premium, and free for new accounts during early access; the core bookmark manager is free. See how it works or the full feature list.
What this checker reads, signal by signal
The checker reads eight kinds of date signals from a single URL, and shows each one it finds in the results table with a plain-English explanation.
JSON-LD dateModified and datePublished
JSON-LD is structured data embedded in a <script type="application/ld+json"> block, mainly for search engines. WordPress SEO plugins, Shopify themes, news CMSs and documentation generators fill dateModified from the post’s own edit timestamp, which makes it the strongest signal the checker can find. datePublished appears as its own row: it tells you how old the page is, not how fresh.
Meta tags
The checker reads article:modified_time, article:published_time, og:updated_time, date and dc.date from the page’s <head>. These usually come from the same source as JSON-LD, so when both exist and agree, confidence is high.
Visible dates
A <time datetime="…"> element near the top of the page usually marks the byline date. It’s what readers see, so it’s weighted below structured data but above technical signals.
Sitemap lastmod
The checker reads robots.txt, follows its sitemap links, and searches for the exact URL, with limits on how many files it opens and how large they can be. Very large sitemaps may be skipped, so a missing sitemap row doesn’t prove the site has no sitemap.
HTTP Last-Modified
The server’s own header, read on the same request as the page. When the value is within a few minutes of the check, the checker marks it “time of request” and leaves it out of the best guess.
Wayback first and latest capture
The oldest and newest Internet Archive snapshots of the URL. They bracket the page’s known life: it existed by the first capture and still existed at the latest one. For a list of URLs, the Bulk Wayback Machine Checker gives the same first and last capture for hundreds of links at once.
How the best guess is picked
The best guess follows a fixed priority: JSON-LD modified date, then meta modified tags, then the sitemap lastmod, then the HTTP Last-Modified header (only if it isn’t simply the time of the request), then the latest Wayback capture. The first signal found in that order becomes the best guess, and its source sets the confidence badge. If signals disagree by a year or more, a warning appears so you know to look closer.
Privacy and limits
The checker fetches pages from Bookmend’s server, so the site you check sees a request from that server, identified as BookmendToolsBot, not from your browser. The URL you enter is sent to the checker to run the lookup; nothing else about you is attached to it. It checks one URL at a time, refuses private and local network addresses, and applies a light rate limit so the service can stay free.
It can’t read pages that require a login, pages behind hard paywalls, or sites that block automated visitors. It can’t recover deleted edit history from a site’s database, and it can’t tell you who made a change. When a site hides its dates, the most honest answer is a range from the Wayback Machine, and the checker says so rather than inventing a precise date. Every answer to “when was website last updated” comes with its sources.
Written by Nafiul Hasan, Founder, Bookmend. Last updated.