Available for new work Remote - worldwide hello@wordpressfixe.com
wordpressfixe

Troubleshooting

WordPress white screen of death: what it means and how to fix it

The white screen is not one bug. It is WordPress hiding a PHP fatal error from visitors — and the fix starts with making that error visible again.

The white screen of death is the most misread failure in WordPress. Visitors see a blank page. The admin may see the same. There is no error message, no stack trace, and often no clue in the browser console. That is not because nothing is wrong — it is because PHP is configured to hide errors from the public, and a fatal error stops execution before WordPress can render anything.

So the first step is not “disable all plugins”. It is making the error visible. Once you know which file and line threw the fatal, the fix usually takes minutes.

What a white screen actually means

A PHP fatal error halts the request. If WP_DEBUG_DISPLAY is off and display_errors is disabled at the server level, PHP returns an empty 200 or 500 response instead of the message. Common triggers:

  • A plugin calling a function that no longer exists after an update
  • A theme written for an older PHP version running on 8.2 or 8.3
  • Exhausting the memory_limit on a heavy page or in the admin
  • A corrupted core file from a partial update or a bad upload
  • A conflict where two plugins declare the same class or function
  • An uploaded file with a syntax error — one missing semicolon is enough

Step 1: get the real error

You need the PHP error log. There are three places to look, in this order:

  1. The hosting panel. Most cPanel, Plesk and GridPane installs expose an error log, usually named error_log in the account root or the domain directory.
  2. A debug log in WordPress. Add these lines to wp-config.php, above the “stop editing” comment:
define('WP_DEBUG', true);
define('WP_DEBUG_LOG', true);
define('WP_DEBUG_DISPLAY', false);

Reload the broken page, then open wp-content/debug.log. The last entry is your fault. It looks like this:

PHP Fatal error: Uncaught Error: Call to undefined function
acme_helper() in /wp-content/plugins/acme-blocks/inc/render.php:84

That single line tells you the plugin, the file and the line. Everything else is confirmation.

Step 2: disable the offending component safely

If the admin is also blank, you cannot click “deactivate”. Do it at the file level with SFTP, or with WP-CLI if you have SSH:

wp plugin deactivate acme-blocks --skip-plugins --skip-themes

No WP-CLI? Rename the plugin folder over SFTP:

/wp-content/plugins/acme-blocks/
→ /wp-content/plugins/acme-blocks.off/

WordPress will deactivate it on the next request. If renaming the folder fixes the site, you have your component. Do the same with the active theme if the error points at wp-content/themes/.

Step 3: if there is no error log

Some hosts disable logging entirely. In that case, bisect instead of guessing:

  1. Rename plugins to plugins.off and create an empty plugins folder. If the site returns, it is a plugin.
  2. Move half the plugin folders back. Reload. Narrow the half that breaks.
  3. Repeat until one folder remains.
  4. Check that plugin’s readme.txt for a “Requires PHP” line and compare it with your host.

This takes four or five requests and no database access. It is slower than reading a log, but it is reliable.

Step 4: rule out memory and the server

If nothing is logged and disabling plugins changes nothing, raise the memory limit temporarily. Add to wp-config.php:

define('WP_MEMORY_LIMIT', '256M');
define('WP_MAX_MEMORY_LIMIT', '512M');

Also check .htaccess for a stray character. A malformed directive can produce a 500 before WordPress loads at all — which looks identical in the browser. If a plain static HTML file in the web root loads but every PHP request fails, the fault is above WordPress: PHP version, handler configuration, or file permissions.

Step 5: fix the cause, not the symptom

Deactivating the plugin restores the site and leaves you without the feature. The real decision is one of four:

  • Update the plugin, if a compatible release exists.
  • Pin the previous version, if the regression is recent and a fix is coming.
  • Replace it, if the project is abandoned — check the last update date, not the download count.
  • Patch a small compatibility issue in a child plugin or a must-use plugin, if the change is genuinely tiny.

Do not edit the plugin directly. The next update overwrites your change and you will be back here with no memory of why.

What not to do

  • Do not reinstall WordPress core first. It rarely causes a white screen and it overwrites nothing useful.
  • Do not delete plugins to “clean up”. You lose the diagnosis and may break features that were working.
  • Do not restore a backup as the first move. You also restore the vulnerability or bug, and you lose everything created since.
  • Do not leave WP_DEBUG_DISPLAY on in production. It prints file paths to visitors.

When to bring someone in

If the site is a business asset and has been down for more than an hour, the cost of a careful diagnosis is almost always lower than the cost of guessing. A fix that starts with the error log typically takes twenty minutes of investigation; the same fault can eat an afternoon when the log is ignored.

If you would rather hand it over, emergency WordPress fixes covers exactly this: error log first, root cause, site back online, and a short written summary of what changed.

Written by

Sahed Rana Neka

PHP and WordPress work focused on repair: emergency fixes, malware cleanup, performance and migrations. Portfolio on sahed.dev.

Need this fixed rather than explained?

MoreKeep reading

Other notes from the same work.