The State of Web Vitals

FCP distribution

How FCP is distributed across real-user data, plus pass-rate breakdown.

Field data PhoneDesktopAll Scope All sites Q2 2026 edition · Phone field outcomes
1

At a glance the headline numbers for FCP distribution

How FCP is distributed across real-user data, plus pass-rate breakdown.

75.5%
of sites good on FCP
1.2s
Typical site
2.6s
Worst 10% of sites

75.5% of sites good on FCP. The typical site's FCP is 1.2s. The worst 10% are above 2.6s.

The State of Web Vitals · Q2 2026 · 189,915 sites · phone field datacorewebvitals.io/state-of-cwv
2

The FCP distribution site count at each FCP, from good to poor

0 5,454 10,908 16,361 21,815
0ms–250msms17 sites<0.1% of all sites
250ms–500msms2,550 sites2.2% of all sites
500ms–750msms13,967 sites12.2% of all sites
750ms–1sms21,815 sites19% of all sites
1s–1.3sms20,149 sites17.6% of all sites
1.3s–1.5sms15,169 sites13.2% of all sites
1.5s–1.8sms11,016 sites9.6% of all sites
1.8s–2sms7,606 sites6.6% of all sites
2s–2.3sms5,537 sites4.8% of all sites
2.3s–2.5sms4,221 sites3.7% of all sites
2.5s–2.8sms2,925 sites2.6% of all sites
2.8s–3sms2,204 sites1.9% of all sites
3s–3.3sms1,753 sites1.5% of all sites
3.3s–3.5sms1,245 sites1.1% of all sites
3.5s–3.8sms952 sites0.8% of all sites
3.8s–4sms730 sites0.6% of all sites
4s–4.3sms544 sites0.5% of all sites
4.3s–4.5sms422 sites0.4% of all sites
4.5s–4.8sms337 sites0.3% of all sites
4.8s–5sms252 sites0.2% of all sites
5s–5.3sms211 sites0.2% of all sites
5.3s–5.5sms155 sites0.1% of all sites
5.5s–5.8sms131 sites0.1% of all sites
5.8s–6sms100 sites0.1% of all sites
6s–6.3sms88 sites0.1% of all sites
6.3s–6.5sms70 sites0.1% of all sites
6.5s–6.8sms62 sites0.1% of all sites
6.8s–7sms51 sites<0.1% of all sites
7s–7.3sms42 sites<0.1% of all sites
7.3s–7.5sms24 sites<0.1% of all sites
7.5s–7.8sms16 sites<0.1% of all sites
7.8s–8sms17 sites<0.1% of all sites
8sms and up148 sites0.1% of all sites
p50 = 1.2s
p75 = 1.8s
p90 = 2.6s
p99 = 5s
0ms–250ms 250ms–500ms 500ms–750ms 750ms–1s 1s–1.3s 1.3s–1.5s 1.5s–1.8s 1.8s–2s 2s–2.3s 2.3s–2.5s 2.5s–2.8s 2.8s–3s 3s–3.3s 3.3s–3.5s 3.5s–3.8s 3.8s–4s 4s–4.3s 4.3s–4.5s 4.5s–4.8s 4.8s–5s 5s–5.3s 5.3s–5.5s 5.5s–5.8s 5.8s–6s 6s–6.3s 6.3s–6.5s 6.5s–6.8s 6.8s–7s 7s–7.3s 7.3s–7.5s 7.5s–7.8s 7.8s–8s 8s+
Good (≤1.8s) Needs improvement Poor (>3s) Percentile markers Total: 114,526 sites

The tallest bar is between 750ms and 1s. 19% of sites are in that one range, 800ms under Google's 1.8s limit. Half of all sites are at 1.2s or less. The tail is long. The worst 10% are above 2.6s and the worst 1% above 5.0s. That is 4 times the typical site.

The State of Web Vitals · Q2 2026 · 189,915 sites · phone field datacorewebvitals.io/state-of-cwv
3

The FCP pass rate the share of sites that are good, needs improvement and poor

FCP
75.5%
18.1%
6.4%
Good Needs Improvement Poor

75.5% of sites are good on FCP. 18.1% are in the needs improvement band, between 1.8s and 3s. 6.4% are poor, above 3s. A miss is usually needs improvement, not poor.

The State of Web Vitals · Q2 2026 · 189,915 sites · phone field datacorewebvitals.io/state-of-cwv
4

Why this matters for the Core Web Vitals, and where to start fixing it

First Contentful Paint is when the first piece of content appears, the first text or image that tells the visitor something is happening. It is a diagnostic metric rather than a Core Web Vital, but it sits on the path to LCP: if the first paint is slow, the largest paint is almost always slow too. FCP is usually held up by the server response and by render-blocking CSS and JavaScript in the head.

Start with what blocks the first paint. Cut render-blocking resources, remove the CSS you do not use, and get the critical bytes to the browser as early as you can. Early Hints can let the browser start fetching key resources before the HTML even arrives. Improving FCP tends to pull LCP along with it.

How are sites doing on FCP?

75.5% of sites have a good FCP. The typical site sits at 1.8s at the 75th percentile; the slowest 1% pass 5.0s.

Go fix it Eliminate render-blocking resources → Remove unused CSS → 103 Early Hints →
Related signals CLS → LCP → TTFB → INP → Chrome field data from 189,915 sites, representing millions of real page loads · How we measured
Live queries (2) — admin only
Query #1: 80.2 ms 
SELECT COUNT(*) AS count,
            COUNT(*) FILTER (WHERE crux."phone".fcp <= 1800) * 1.0 / COUNT(*) AS good,
            COUNT(*) FILTER (WHERE crux."phone".fcp > 1800 AND crux."phone".fcp <= 3000) * 1.0 / COUNT(*) AS needs_improvement,
            COUNT(*) FILTER (WHERE crux."phone".fcp > 3000) * 1.0 / COUNT(*) AS poor,
            quantile_disc(crux."phone".fcp, 0.10) AS p10,
            quantile_disc(crux."phone".fcp, 0.25) AS p25,
            quantile_disc(crux."phone".fcp, 0.50) AS p50,
            quantile_disc(crux."phone".fcp, 0.75) AS p75,
            quantile_disc(crux."phone".fcp, 0.90) AS p90,
            quantile_disc(crux."phone".fcp, 0.99) AS p99
            FROM sites WHERE 1=1 AND crux."phone".fcp IS NOT NULL;
Query #2: 61.4 ms 
SELECT CASE WHEN crux."phone".fcp >= 0 AND crux."phone".fcp < 250 THEN 0 WHEN crux."phone".fcp >= 250 AND crux."phone".fcp < 500 THEN 1 WHEN crux."phone".fcp >= 500 AND crux."phone".fcp < 750 THEN 2 WHEN crux."phone".fcp >= 750 AND crux."phone".fcp < 1000 THEN 3 WHEN crux."phone".fcp >= 1000 AND crux."phone".fcp < 1250 THEN 4 WHEN crux."phone".fcp >= 1250 AND crux."phone".fcp < 1500 THEN 5 WHEN crux."phone".fcp >= 1500 AND crux."phone".fcp < 1750 THEN 6 WHEN crux."phone".fcp >= 1750 AND crux."phone".fcp < 2000 THEN 7 WHEN crux."phone".fcp >= 2000 AND crux."phone".fcp < 2250 THEN 8 WHEN crux."phone".fcp >= 2250 AND crux."phone".fcp < 2500 THEN 9 WHEN crux."phone".fcp >= 2500 AND crux."phone".fcp < 2750 THEN 10 WHEN crux."phone".fcp >= 2750 AND crux."phone".fcp < 3000 THEN 11 WHEN crux."phone".fcp >= 3000 AND crux."phone".fcp < 3250 THEN 12 WHEN crux."phone".fcp >= 3250 AND crux."phone".fcp < 3500 THEN 13 WHEN crux."phone".fcp >= 3500 AND crux."phone".fcp < 3750 THEN 14 WHEN crux."phone".fcp >= 3750 AND crux."phone".fcp < 4000 THEN 15 WHEN crux."phone".fcp >= 4000 AND crux."phone".fcp < 4250 THEN 16 WHEN crux."phone".fcp >= 4250 AND crux."phone".fcp < 4500 THEN 17 WHEN crux."phone".fcp >= 4500 AND crux."phone".fcp < 4750 THEN 18 WHEN crux."phone".fcp >= 4750 AND crux."phone".fcp < 5000 THEN 19 WHEN crux."phone".fcp >= 5000 AND crux."phone".fcp < 5250 THEN 20 WHEN crux."phone".fcp >= 5250 AND crux."phone".fcp < 5500 THEN 21 WHEN crux."phone".fcp >= 5500 AND crux."phone".fcp < 5750 THEN 22 WHEN crux."phone".fcp >= 5750 AND crux."phone".fcp < 6000 THEN 23 WHEN crux."phone".fcp >= 6000 AND crux."phone".fcp < 6250 THEN 24 WHEN crux."phone".fcp >= 6250 AND crux."phone".fcp < 6500 THEN 25 WHEN crux."phone".fcp >= 6500 AND crux."phone".fcp < 6750 THEN 26 WHEN crux."phone".fcp >= 6750 AND crux."phone".fcp < 7000 THEN 27 WHEN crux."phone".fcp >= 7000 AND crux."phone".fcp < 7250 THEN 28 WHEN crux."phone".fcp >= 7250 AND crux."phone".fcp < 7500 THEN 29 WHEN crux."phone".fcp >= 7500 AND crux."phone".fcp < 7750 THEN 30 WHEN crux."phone".fcp >= 7750 AND crux."phone".fcp < 8000 THEN 31 WHEN crux."phone".fcp >= 8000 THEN 32 END AS bucket_idx, COUNT(*) AS count FROM sites WHERE 1=1 AND crux."phone".fcp IS NOT NULL GROUP BY bucket_idx;
JSON file lookups (4)
20260625/meta.json 0.03 ms
20260625/types.json 0.73 ms
20260625/loaf-scripts.json 0.53 ms
20260625/menu.json 0.17 ms