The State of Web Vitals

WordPress unused CSS bytes and the Core Web Vitals

38.1% of audited sites (24,090) parse over 600 KB of unused CSS rules before first paint, passing LCP at 60% (2.2s median) compared with 86% (1.5s median) under 140 KB.

Field data PhoneDesktopAll Scope All sites October 2026 edition · All devices field outcomes · WordPress sites only
WordPress Scoped to 63,295 sites out of 203,892 analyzed (cms) Compare with all sites →

10.2% of WordPress sites use under 140 KB of unused CSS unused CSS bytes, and 86% of those sites have a good LCP, against 70% of all WordPress sites.

The State of Web Vitals · October 2026 · WordPress: 63,295 crawled sites · 60,855 with all devices field data · 63,268 sites with unused CSS bytesBy Arjen Karel · corewebvitals.io/state-of-cwv
1

LCP pass rate

Sites with under 140 KB of unused CSS unused CSS bytes have a good LCP most often: 86%, against 70% of all WordPress sites. Sites with over 600 KB of unused CSS unused CSS have it least often, at 60%.

SitesGood, needs improvement, poor LCPMedian
All WordPress sites60,85570%1.9s
Under 140 KB unused 10.2%6,20086%1.5s
140 to 300 KB unused 21.9%13,23079%1.7s
300 to 600 KB unused 29.8%18,12772%1.9s
Over 600 KB unused 38.1%23,27160%2.2s
GoodNeeds improvementPoorSites: with Chrome field data
The State of Web Vitals · October 2026 · WordPress: 63,295 crawled sites · 60,855 with all devices field data · 63,268 sites with unused CSS bytesBy Arjen Karel · corewebvitals.io/state-of-cwv
2

Unused CSS bytes usage

Over 600 KB of unused CSS is the most used: 38.1% of WordPress sites. Under 140 KB of unused CSS is 10.2% of WordPress sites.

Over 600 KB unused 38.1% 300 to 600 KB unused 29.8% 140 to 300 KB unused 21.9% Under 140 KB unused 10.2%
Over 600 KB unused 38.1% 300 to 600 KB unused 29.8% 140 to 300 KB unused 21.9% Under 140 KB unused 10.2%
The State of Web Vitals · October 2026 · WordPress: 63,295 crawled sites · 60,855 with all devices field data · 63,268 sites with unused CSS bytesBy Arjen Karel · corewebvitals.io/state-of-cwv
3

The four LCP phases

Across the four LCP sub-parts in Chrome's field data, the largest gap between under 140 KB of unused CSS and over 600 KB of unused CSS sits in time to first byte (TTFB): 843 ms median for under 140 KB of unused CSS against 1296 ms for over 600 KB of unused CSS. Overall median field LCP is 1.5s for under 140 KB of unused CSS (TTFB 843 ms, load delay 214 ms, load duration 156 ms, render delay 212 ms) versus 2.2s for over 600 KB of unused CSS.

Median LCP phase breakdown (Chrome field p75) Load delay Median LCP
WordPress sites with unused CSS bytes 43,6061127 ms340 ms340 mslab 340 ms1.9s
Under 140 KB unused 4,163843 ms214 mslab 214 ms1.5s
140 to 300 KB unused 9,2841007 ms279 ms279 mslab 279 ms1.7s
300 to 600 KB unused 13,1161117 ms325 ms325 mslab 325 ms1.9s
Over 600 KB unused 17,0431296 ms433 ms261 ms433 mslab 433 ms2.2s
TTFB Resource load delay Resource load time Element render delay
The State of Web Vitals · October 2026 · WordPress: 63,295 crawled sites · 60,855 with all devices field data · 63,268 sites with unused CSS bytesBy Arjen Karel · corewebvitals.io/state-of-cwv
4

Unused CSS bytes per framework

On jQuery, 40% of sites use over 600 KB of unused CSS unused CSS bytes, and 69% of jQuery sites have a good LCP, against 70% of all sites. On Vue, 48% of sites use over 600 KB of unused CSS, and 68% of Vue sites have a good LCP.

FrameworkUnused CSS bytes on its sitesGood LCP
WordPress sites with unused CSS bytes 63,268 sitesOver 600 KB unused 38%300 to 600 KB unused 30%140 to 300 KB unused 22%Under 140 KB unused 10%70%
Alpine 380 sitesOver 600 KB unused 28%Under 140 KB unused 24%300 to 600 KB unused 24%140 to 300 KB unused 23%80%
jQuery 60,290 sitesOver 600 KB unused 40%300 to 600 KB unused 30%140 to 300 KB unused 22%69%
Vue 4,844 sitesOver 600 KB unused 48%300 to 600 KB unused 32%140 to 300 KB unused 13%68%
React 2,233 sitesOver 600 KB unused 52%300 to 600 KB unused 30%140 to 300 KB unused 12%61%
Over 600 KB unused 300 to 600 KB unused 140 to 300 KB unused Under 140 KB unused
The State of Web Vitals · October 2026 · WordPress: 63,295 crawled sites · 60,855 with all devices field data · 63,268 sites with unused CSS bytesBy Arjen Karel · corewebvitals.io/state-of-cwv
5

Unused CSS bytes or the framework

Does the framework explain the unused CSS bytes difference?

On jQuery, 85% of sites with under 140 KB of unused CSS unused CSS bytes have a good LCP, against 60% of jQuery sites with over 600 KB of unused CSS. Within jQuery, sites with under 140 KB of unused CSS have a good LCP more often. On Vue, 78% of sites with under 140 KB of unused CSS unused CSS bytes have a good LCP, against 61% of Vue sites with over 600 KB of unused CSS. Within Vue, sites with under 140 KB of unused CSS have a good LCP more often.

Why unused CSS slows both network transfer and style calculation

Browsers must match every selector in the CSSOM against the DOM tree during layout and style recalculation. Across 6,473 sites (10.2%) with under 140 KB of unused CSS, 86% pass field LCP at a 1.5s median, whereas the 38.1% (24,090) over 600 KB drop to 60% good LCP (2.2s median).

The State of Web Vitals · October 2026 · WordPress: 63,295 crawled sites · 60,855 with all devices field data · 63,268 sites with unused CSS bytesBy Arjen Karel · corewebvitals.io/state-of-cwv
6

Understanding unused CSS bytes

Strip unused CSS rules per template or switch to block themes that enqueue styles only for rendered blocks. On WordPress, 10.2% of sites keep unused CSS under 140 KB (86% good LCP) while 38.1% exceed 600 KB (60% good LCP).

Go fix it Remove unused CSS → Eliminate render-blocking resources →
The State of Web Vitals · October 2026 · WordPress: 63,295 crawled sites · 60,855 with all devices field data · 63,268 sites with unused CSS bytesBy Arjen Karel · corewebvitals.io/state-of-cwv
Related signals Iframes per page → Image fetchpriority → Responsive image markup → Script coverage (used vs unused) → By Arjen Karel · 63,295 crawled WordPress sites (60,855 with all devices Chrome field data, 63,268 sites with unused CSS bytes, October 2026) · How we measured