Largest Contentful Paint (LCP): What It Measures and How to Fix It
Largest Contentful Paint is the moment the biggest image or text block in the first screen appears. It is the Core Web Vital that answers the question visitors actually ask: has the page loaded yet? Google wants it at 2.5 seconds or less for three in four visits. Here is how LCP works, what slows it down and how to fix it for a whole template, not just one page.
What Is Largest Contentful Paint?
Largest Contentful Paint (LCP) reports the time, from when the visitor starts loading the page, until the largest image, video or text block visible in the viewport is rendered. It is one of the three Core Web Vitals, next to Interaction to Next Paint and Cumulative Layout Shift.
The largest element can change while the page loads. A heading may be the largest thing at first, then a hero image takes over once it arrives. The browser keeps reporting new candidates and stops as soon as the visitor taps, scrolls or presses a key, so the last candidate before that is the LCP.
LCP replaced older load metrics such as onload because it tracks what people see, not when the last script finished. SEO ProCheck keeps a short LCP glossary entry if you need the one-line version for a client.
LCP Thresholds at a Glance
| Rating | LCP at the 75th percentile |
|---|---|
| ✓ Good | ≤ 2.5 s |
| ◐ Needs improvement | 2.5 to 4.0 s |
| – Poor | > 4.0 s |
The numbers come from web.dev. A page passes when the 75th percentile of real visits is good, measured separately for phones and desktops. That means one slow visit in four is allowed; two are not.
| Number | Value | Source |
|---|---|---|
| Good line | 2.5 s | web.dev |
| Poor line | 4.0 s | web.dev |
| Percentile used | 75th | web.dev |
| Field data window | 28 days | CrUX |
| Ideal TTFB share | about 40% | Optimize LCP |
Which Elements Can Be the LCP Element
Not everything on the page is a candidate. The LCP definition counts these element types:
| Element | Counts for LCP | Note |
|---|---|---|
| img | ✓ | Most common LCP element |
| image inside svg | ✓ | |
| video | ✓ | Poster or first frame |
| CSS background with url() | ✓ | Found late by the browser |
| CSS gradient | – | |
| Block of text | ✓ | Headings, paragraphs |
| Content outside the viewport | – | Only the visible part counts |
On most templates the LCP element is a hero image, a product photo or the first heading. Once you know which one it is for each template, you know what to fix. Chrome DevTools shows it in the Performance panel, and PageSpeed Insights names it under the LCP breakdown.
The Four Parts of LCP
web.dev's guide to optimizing LCP splits the time into four parts. Measure each one and you know which fix to make.
| Part | From | To | Target share |
|---|---|---|---|
| Time to first byte | Navigation start | First byte of HTML | about 40% |
| Resource load delay | First byte | Image request starts | under 10% |
| Resource load duration | Request starts | Image fully loaded | about 40% |
| Element render delay | Image loaded | Element painted | under 10% |
The two short parts are where most of the easy wins are. A long load delay means the browser found the image late: it was added by JavaScript, set as a CSS background, or lazy-loaded. A long render delay means something blocked the paint after the image had arrived, usually a render-blocking style sheet or a script. For text LCP elements there is no resource to load, so only the first byte and the render delay matter.
Common Causes of a Slow LCP and Their Fixes
| Cause | Part it hurts | Fix |
|---|---|---|
| Slow server or no cache | TTFB | Cache HTML, use a CDN |
| Redirect before the page | TTFB | Link the final URL |
| LCP image added by JavaScript | Load delay | Put it in the HTML |
| LCP image as CSS background | Load delay | Use img, or preload it |
| loading=lazy on the LCP image | Load delay | Remove it |
| No fetchpriority=high | Load delay | Add fetchpriority=high |
| Image far too large | Load duration | Resize, use srcset |
| Old format, poor compression | Load duration | WebP or AVIF |
| Image on another host | Load duration | Same host, or preconnect |
| Render-blocking CSS or scripts | Render delay | Inline critical CSS, defer scripts |
| Web font hides the text | Render delay | font-display: swap |
| Fade-in animation on the hero | Render delay | Show it at once |
The last row is easy to miss. A hero that fades in over a second counts the end of the fade, not the start. SEO ProCheck describes a site that improved LCP by removing image transitions.
Fixing the LCP Image, Step by Step
- Find the LCP element for each template in PageSpeed Insights or DevTools.
- Put it in the initial HTML as an img with a src, not a placeholder that a script fills in later.
- Tell the browser it matters with fetchpriority=high, and never loading=lazy.
- Serve the right size with srcset and sizes, in WebP or AVIF.
- Set width and height so the space is held, which also protects CLS.
- Check the first byte. If TTFB alone is over 0.8 seconds, the server is the problem, not the image.
<img src="/img/hero-1200.webp"
srcset="/img/hero-800.webp 800w, /img/hero-1200.webp 1200w"
sizes="100vw" width="1200" height="600"
fetchpriority="high" alt="Trail running shoes on a rocky path">
If the hero has to stay a CSS background, preload it so the browser does not wait for the style sheet:
<link rel="preload" as="image" href="/img/hero-1200.webp" fetchpriority="high">
The fetch priority article on web.dev explains how the browser orders requests. Image formats and sizes are covered in depth in the image optimization SEO guide.
How to Measure LCP: Field and Lab
| Tool | Field | Lab | Shows the LCP element |
|---|---|---|---|
| PageSpeed Insights | ✓ | ✓ | ✓ |
| Search Console Core Web Vitals report | ✓ | – | – |
| Chrome DevTools Performance panel | – | ✓ | ✓ |
| Chrome UX Report API | ✓ | – | – |
| web-vitals library | ✓ | – | ✓ with attribution |
Field data is what Google uses; lab data is what tells you why. Lab LCP on a throttled phone is often slower than the field number for repeat visitors with a warm cache, and faster for people on poor networks. Read both, decide with the field number, and fix with the lab trace. The Core Web Vitals checker guide compares the tools in more detail, and the bulk page speed test guide shows how to measure every template at once.
Find LCP Problems With Neova
Neova reads LCP both ways. With your free Chrome UX Report API key, Real-User Core Web Vitals shows the 75th percentile LCP of real Chrome users on phones, per page group, with a 40-week trend and alerts when the grade changes. With Lighthouse turned on in New Crawl, it runs lab checks on one HTML page per page group, so every template gets a diagnosis. The checks that point at LCP:

- HTML Is Over 2 MB or Core Web Vitals Miss the Good Range, which flags field LCP over 2.5 s
- LCP Image Is Not Discoverable, Not Prioritized or Lazy-Loaded
- LCP Image Is Lazy-Loaded
- Server Response Time Is Slow in Lab or Field Data, for the first part of LCP
- Render-Blocking Requests Delay the First Paint
- Images Are Larger Than Their Displayed Dimensions and Images Are Not Served in a Modern Format
- First Contentful Paint Is Slow
- Content Image Is Delivered as a CSS Background, with the AI reader
Every check runs on every plan, the Free plan included, on up to 1,000 URLs a crawl, on
Questions
What is a good largest contentful paint?
2.5 seconds or less at the 75th percentile of real visits, measured separately for phones and desktops. Between 2.5 and 4 seconds needs improvement, and over 4 seconds is poor.
What is the LCP element?
The largest image, video poster, background image or block of text visible in the viewport before the visitor first interacts. PageSpeed Insights and Chrome DevTools both name it for a given page.
Should I lazy-load the LCP image?
No. loading=lazy makes the browser wait until layout before it requests the image, which adds load delay. Lazy-load images further down the page and give the LCP image fetchpriority=high instead.
Why is my LCP good in Lighthouse but poor in Search Console?
Lighthouse runs one load on a simulated phone; Search Console reports 28 days of real Chrome users on their own devices and networks. Real visitors may land on slower pages, with cold caches or on weak connections. SEO ProCheck's LCP too slow diagnosis page walks through the usual reasons.
Does LCP affect rankings?
Core Web Vitals are part of the page experience signals Google's ranking systems use, but relevance comes first. Good LCP helps most between pages that are otherwise close, and it keeps more visitors on the page.