Guide
Wayback Machine to WordPress: how to restore a site step by step
You can restore a site from the Wayback Machine into WordPress if the archive saved its pages. The work has six steps: check what the archive holds, pick the snapshot window, download the original files, clean them, rebuild the site in WordPress on the same URLs, and decide what every missing address should answer. This guide walks through each step, with the commands and checks I use myself.
By Dmitriy Gabdrakhmanov · Updated 9 October 2026
What the Wayback Machine saves, and what it does not
The Wayback Machine saves what a crawler saw from the outside: HTML pages, stylesheets, scripts and images, each captured on its own date. It does not save anything that lived only on the server.
Comes back from the archive:
- Page text and layout as HTML
- Images, stylesheets and scripts that were crawled
- Titles, meta descriptions and internal links
- Redirects the site answered with
Does not come back:
- The database itself: users, passwords, orders
- Comments and pages that were never crawled
- Contact forms, search and anything else run by PHP
- Files blocked by robots.txt or removed from the archive on request
Two things surprise most people. First, every URL has its own capture dates, so a restored site is assembled from snapshots taken weeks or years apart. Second, a page that looks complete in the archive viewer may be missing images or styles that were never captured. Plan for gaps from the start.
The full list of what the archive cannot restore →
Step 1: check what the archive holds
Before downloading anything, list every URL the archive has for the domain. The Wayback Machine has a public index called the CDX API. One request returns every captured address with its status and date:
https://web.archive.org/cdx/search/cdx?url=example.com/*&output=json&fl=original,statuscode,timestamp&collapse=urlkey
Replace example.com with the domain. collapse=urlkey keeps one row per address. Add &filter=statuscode:200 to see only pages that answered normally. What to look for:
- How many HTML pages answered 200. This is the real size of the site you can restore.
- Which addresses matter. If you have a backlink export from your SEO tool, compare it with this list. Every URL with links pointing to it should exist in the archive.
- Parameter addresses. Old WordPress sites often have links to
?p=123,?page_id=45or?cat=5. Note them now: they need special handling in step 4.
If the list is short or full of redirects, the archive may not hold enough for a restoration. That is worth knowing before you spend money or time.
Step 2: pick the snapshot window
A domain often changes hands. After it expires, the archive may capture a parking page, a spam site or a different business on the same domain. Restoring those captures brings back the wrong site.
- Open the captures by year in the archive calendar and find when the original site stopped.
- Set the window to end before that date: the last good year of the original site.
- Check a few deep pages, not only the homepage. Spam often appears on new URLs while the homepage still looks fine.
Owning the domain does not give you rights to a previous owner’s content. If the site was not yours, check the rights before you publish the restored pages.
Step 3: download the original files
The archive viewer adds its own toolbar and rewrites every link to web.archive.org. To get the page as it was served, add id_ after the timestamp in the capture address:
https://web.archive.org/web/20190315000000id_/https://example.com/about/
This returns the original bytes of the capture, without the toolbar and without rewritten links. Do the same for stylesheets, scripts and images. A few rules keep the download working:
- Go slowly. The archive throttles fast clients. If it answers
429, wait the time it asks for. Pushing on can get your IP blocked for an hour or more. About one request per second is safe. - Keep a log. Write down every URL, its status and the capture you used. You will need it in steps 5 and 6.
- Expect missing assets. When an image was never captured in your window, look for the nearest capture outside it before giving up.
Step 4: rebuild in WordPress on the original URLs
This is the step that decides whether the restoration was worth it. Backlinks, bookmarks and search results point to the old addresses. If the new site answers on different URLs, those links lead nowhere.
Keep every address exactly
Use the same paths, with or without the trailing slash as the original had, and the same file names such as /page.html or /index.php?id=7. Do not “clean up” old URLs into pretty permalinks: a renamed URL is a new URL.
Parameter addresses need extra care
By default WordPress answers ?p=123 with a redirect to the pretty permalink of post 123, and only if a post with that ID exists. If the old site was linked as ?p=123, the restored site has to answer that exact address with the right page. That means either matching post IDs and permalink settings to the original, or a small router that maps old parameter addresses to restored pages.
Pages as HTML blocks or as editor blocks
The fastest faithful route is to put each page’s original markup into WordPress as an HTML block, with the original stylesheet in the theme. The page looks and reads as it did, and the text stays editable. Rebuilding every page in visual editor blocks gives easier editing later, but it is a redesign job and takes far longer.
Step 5: decide what every missing address answers
Some addresses will not come back: the archive never captured them, or only an error page was saved. Each of them needs a deliberate answer.
| The address | Answer with | Why |
|---|---|---|
| Restored page | Answer with200 | WhyThe page is back with its original content |
| The original site already redirected it | Answer with301 to the same target | WhyKeeps the redirect the site itself had |
| Gone: not in the archive, or not restorable | Answer with410 | WhySays honestly that the page is gone on purpose |
| Anything else that never existed | Answer with404 | WhyNormal “not found” |
A common shortcut is sending every missing address to the homepage. Google treats redirects to an irrelevant page as soft 404s, so those redirects pass nothing useful and hide real gaps. Keep a file with one row per old address: its status, its target and the reason. It doubles as documentation for whoever runs the site next.
How to choose between 301 and 410, address by address →
Step 6: check before launch
Run these checks on the restored site before you point the domain at it:
- Request every URL from your list and your backlink export. Each should answer
200,301or410as planned, never a stray404or a redirect to the homepage. - Search the site files for
web.archive.org, the archive toolbar and dead analytics or ad snippets. None should remain. - Check that each page’s canonical address is the page itself, on the original URL.
- Create
sitemap.xmlwith the restored pages and arobots.txtthat points to it. - Open the key pages on a phone. Old layouts sometimes break on small screens; that is how the original looked, but you should know it.
When all of this passes, switch the domain and submit the sitemap in Google Search Console.
How long it takes and what it costs
| Route | Your time | Cost | What you get |
|---|---|---|---|
| By hand, following this guide | Your timeDays for a small site, weeks for a large one | CostYour hours | What you getExactly what you build |
| A download tool | Your timeMinutes to download, then hours to convert into WordPress and check | CostPer file or per site, often a few dollars | What you getA static or PHP copy of the archived pages |
| A restoration service | Your timeMinutes to order | CostA fixed price per site | What you getA working WordPress site, depending on the provider |
The download is the quick part. Steps 4 to 6 take most of the time: keeping the original URLs, deciding what every missing address answers and checking it all. When comparing options, ask what happens to those addresses. The guide Three ways to bring back a site lists six questions to ask any provider.
Questions
Can I restore the original database, users or comments?
No. The archive saved what visitors saw, not what was stored on the server. Users, passwords, orders and comments that were not shown on crawled pages cannot be recovered.
Will my Google rankings come back?
Nobody can promise that. What a restoration can do is make the old addresses answer again with the original content, so links pointing to them reach a real page. How search engines rank the site after that is their decision.
How long does a restoration take?
Downloading takes minutes to hours. A careful rebuild in WordPress with the original URLs and a check of every address takes days for a small site. As a service, I deliver sites up to 2,000 pages in 2–3 working days.
Can I restore a site I do not own?
You can restore archived pages of a domain you own, but owning the domain does not give you rights to the previous owner’s content. Check the rights before you publish.
Do I need the Wayback Machine snapshot URL to start?
No. The domain is enough. The CDX index in step 1 lists every captured address, and the right snapshot window comes from the archive calendar in step 2.
Next step
If you would rather not do steps 3 to 6 yourself, I do them as a service: WordPress on the original URLs, ?p=123 addresses included, a 301 or 410 with a reason for every missing address, and a QA report for every URL. Every order starts with a free audit of what the archive holds.