Most teams know Core Web Vitals affect rankings. Far fewer can tell you the actual numbers they have to hit — or that “good” isn’t a single value but a line you must clear for three out of every four real visitors.

There are three vitals in 2026, each measuring a different part of the experience: how fast the main content loads, how quickly the page responds to input, and how much the layout shifts as it settles. Here are the thresholds that separate good from poor.

Core Web Vitals — 2026 LCP Largest Contentful Paint · Loading ≤ 2.5 s Needs ≤ 4.0 s · Poor > 4.0 s INP Interaction to Next Paint · Responsiveness ≤ 200 ms Needs ≤ 500 ms · Poor > 500 ms CLS Cumulative Layout Shift · Stability ≤ 0.10 Needs ≤ 0.25 · Poor > 0.25 You pass only if 75% of real visits clear the “good” line
The 2026 Core Web Vitals pass/fail thresholds, graded on real-user field data.

LCP — loading

Largest Contentful Paint measures how long the biggest visible element — usually a hero image or headline block — takes to render. Good is 2.5 seconds or less. Slow LCP usually traces back to heavy hero images, slow server response, or render-blocking scripts and fonts.

INP — responsiveness

Interaction to Next Paint replaced First Input Delay, and it’s a tougher test: it looks at how quickly the page visibly responds across all interactions, not just the first. Good is 200 milliseconds or less. Heavy JavaScript tying up the main thread is the usual cause — and it barely shows up in a lab test that never actually clicks anything.

CLS — visual stability

Cumulative Layout Shift scores how much content jumps around while the page loads. Good is 0.10 or less. The classic causes are images and ad slots without reserved dimensions, and late-loading fonts or embeds that shove the page down just as someone goes to tap.

The rule that decides it: the 75th percentile

Here’s the part that catches teams out. Google doesn’t grade your average visitor — it grades you at the 75th percentile of real users over a rolling 28-day window. It isn’t enough for the typical visit to feel fast; three out of four visits have to clear the “good” line. That’s why a great score on your fast laptop can sit beside a failing grade in Search Console.

Watch the trend, not a single test

Because these are field measurements aggregated over weeks, one spot check tells you little. A regression after a deploy shows up as a slow drift, not a red number on a single run. CrUX Tracker monitors your real-user LCP, INP and CLS at the 75th percentile continuously, so you always know which side of these lines you’re on.

Want to see where your site stands against these thresholds? start tracking your field data on the CrUX Tracker page.