If you need to recover website from Wayback Machine archives, start by resetting your expectations. The Internet Archive’s help page, Using the Wayback Machine, says plainly that its “terms of use do not cover backups for the general public”, that it “can’t guarantee that your site has been or will be archived”, and that it “can no longer offer the service to pack up sites that have been lost.” Nobody at the Archive will send you a zip file. What you get is access to the copies its crawlers made, and the job of turning those copies back into a site is yours.
That still makes it the best salvage source most people have after a lost host account or a hacked install. Below is the order of work, what survives, what does not, and where a recovered copy quietly goes wrong.
First, confirm you have the right to recover it
The Archive’s own help page frames recovery around “a site to which you own the rights”. That is the only case it covers. If you are looking at an old site you never owned, perhaps because you just bought its lapsed domain, the archive tells you what was there. It does not hand you a licence to republish someone else’s writing and images.
If the domain itself expired, deal with that before anything else. A recovered site with nowhere to live is a folder of files. Our guide to getting an expired domain back covers the stages, and the Domain History Checker shows a year-by-year Wayback timeline with past page titles, topic changes, gaps and redirects, which tells you quickly which years are worth digging into.
Recover website from Wayback Machine captures: list every file first
The help page gives one pattern for seeing everything captured under a site: web.archive.org/*/www.example.com/*. Swap in your own domain and you get a list of archived URLs, not just the homepage. Treat that list as your inventory. It is also the start of your redirect map, much like the crawl you would run before a site migration.
For any single URL, the date code in an archived address “translates as yyyymmddhhmmss”. Replace that code with an asterisk and you see every capture date for that one page. Use this when a key page looks wrong: there may be a cleaner capture a few weeks earlier.
Orphan pages are the usual gap. The help page notes that “if there are no links to your pages, the robot won’t find it.” If an old sitemap or analytics export survives, compare it with the archive list. Our post on how to find all pages on a website covers other ways to rebuild that list.
Pick capture dates by colour
The calendar view colour-codes each capture by the response the crawler got. The Archive’s advice is short: “Most of the time you will probably want to select the blue dots or links.”
| Calendar colour | Response | Use it for recovery? |
|---|---|---|
| Blue | 2nn, good | Yes, this is the page as served |
| Green | 3nn, redirect | Only to learn where the URL pointed |
| Orange | 4nn, client error | No, the page was missing then |
| Red | 5nn, server error | No, the server failed then |
Green captures are still useful. They show which old URLs had already been redirected, which matters when you rebuild the redirect map. Our note on 301 vs 302 redirects covers which type to use when you put them back.
Download: by hand or with a third-party service
The Help Center does not describe an official export tool. For a small site, saving pages by hand from blue captures is slow but fully under your control. For larger sites, the Archive’s help page “Can I rebuild my website using the Wayback Machine?” names three third-party services: waybackrebuilder.com, waybackdownloader.com and waybackdownloads.com. It adds a caveat worth repeating: “We do not have direct experience working with any of these sites, so please investigate the services to see whether they will meet your individual needs.” We have not tested them either, so treat this as a list of names, not a recommendation.
Whichever route you take, do not assume the output is complete. That is the next step.
Fix the mixed dates and the missing images
This is where recoveries go wrong without anyone noticing. The help page says “Not every date for every site archived is 100% complete.” When a linked file is missing, the Wayback Machine “will grab the closest available date”, and if it was never archived it “will look for the link on the live web and grab it if available.” A page that looks whole in your browser can be stitched from several dates, and part of it may come from whatever sits on the domain today.
So check the timestamp of every file you pull, especially stylesheets and images. Broken images, the Archive explains, “occur when the images are not available on our servers. Usually this means that we did not archive them.” You can test any one file by entering its URL into the Wayback Machine search box. If it is not there, you need the original from a local copy, an email attachment or a re-shoot.
What will not come back
JavaScript is “often hard to archive”, and “if javascript needs to contact the originating server in order to work, it will fail when archived.” The “Wayback Machine General Information” page goes further: where a page has forms or other elements that need the original host, “the archive will not contain the original site’s functionality.” Pages behind a password are not archived. Databases, logins, contact forms, carts and server code were never captured, so they are a rebuild, not a recovery. If the old site leaned on client-side rendering, our JavaScript SEO post explains why the archive and search engines both saw less than you did. As the help page puts it, “simple html is the easiest to archive.”
Rebuild, redirect and check
Once you have clean files, move the content into a CMS rather than serving the raw archive HTML, which often still carries archive markup and absolute links. Keep the old URL paths where you can. Where you cannot, map each old URL from your inventory to its new home with a 301, and let truly dead pages return a proper error; our guide to 404 errors covers when that is fine. A recovery is also a natural moment for the checks in a website redesign.
Finally, make a real backup. If you want the new site in the archive, Save Page Now saves “a specific page one time” and “does not save multiple pages, directories or entire sites.” There is also “a 3-10 hour lag time” between a crawl and its appearance. Neither is a backup plan.
A limit to keep in mind: none of this tells you how much of your site was captured until you list it. Some sites were “password protected, blocked by robots.txt, or otherwise inaccessible”, and some owners asked to be excluded. If yours was one of them, the archive has nothing to give.
Frequently asked questions
Can the Internet Archive send me a backup of my site?
No. Its help page says its terms of use do not cover backups for the general public and that it can no longer pack up lost sites. You can access archived versions of a site you own the rights to and save them yourself.
How do I see every page the Wayback Machine archived from my site?
Use the pattern web.archive.org/*/www.example.com/* with your own domain. For one URL, replace the date code in its archived address with an asterisk to see all capture dates.
Why are images missing from my recovered pages?
Usually because they were never archived. Check each image URL in the Wayback Machine search box; if it is absent, you need the original from another source.
Can I recover forms, logins or a shop?
No. Anything that needed the original server to work was not captured. Recover the content, then rebuild the functionality.
The takeaway Treat the Wayback Machine as salvage. List the files, choose blue captures, check every timestamp, rebuild what never archived, and redirect the old URLs.
